I've spent a decent amount of the weekend playing FE 0.75; I've been impressed with some things, but others still need more work.  Here's a list of the bugs and balance issues I encountered, starting with the minor ones:

* The fracture spell heals elementals, rather than damaging them.

* The giant form and growth spells have graphics issues when cast on a group -- at first it just enlarges one of the units, though after some fighting the whole group ends up enlarged

* The Forest Drake Lair improvement doesn't have a graphic for building/built, and doesn't do anything once you finish it (even playing as the race with the "serpent pact").  Is the intention that I should be able to recruit forest drakes there?

* While fighting Abeix, he somehow ended up with infinite turns in a row and I had to kill the game and restart the combat.  I think what happened is that I needed to retreat much of my party back in order to let my warriors past the mages, and so he ended up in a position where he had nobody to attack (because nobody was adjacent to the part of the chasm he's on).  Later I tried attacking him again, and determined that it seemed impossible to attack or cast spells on him, so maybe he's just not done yet.

* The "tactical spell cost" discounts stack too well.  I had a character who had 90% in bonuses to tactical spell cost (40% from robe of the archmage, 25% from Path of the Mage, and 25% from the "Affinity" special ability, I think).  So fireball cost 3 mana and flame dart cost 1 and focus cost 0 (because it rounded down) -- it was kinda absurd.  I don't like the idea of "just take the best" bonus, since that makes it pointless to get more similar abilities, but maybe the bonuses should apply multiplicatively (so that character would pay 0.6*0.75*0.7 ~ 30% the normal cost, rather than 5%).

* Fireball is perhaps too powerful against armies -- because an army starts out in a big 3x3 cluster, and it's very difficult to make troops immune to fire until late game, a single character with the fireball spell and the impulsive trait can cast it before the other side moves and wipe out an entire army.  Maybe it should only do the "damage multipled by number of creatures" and not also the "hits a 3x3 cluster of tiles" so that it kills companies but not entire armies?  Or maybe it could do a 2x2 area, and a higher-level spell more like fire storm could provide the "3x3" cluster capability.

* There should be level minimums of becoming e.g. a "fire archmage".  Also, it seems like there should be benefits for a fire archmage for casting lower-level fire spells -- I found myself with a lot of archmage class characters whose most effective thing to do against a single large foe was to go hit them with their sword, not cast spells.

* I ran into an AI player who had a General Carrodus character with about 800 hit points, which seemed to all come from 4 castings of the "Blood Curse" spell (which is a pretty cool idea).  But is this supposed to be possible?

* It seems like you should give your mages daggers rather than staves because the initiative bonuses are better and initiative is all that matters for a mage -- is there a way to make this not the right plan?  One thought for how this could work is that the initiative modified for a weapon only applies when you used the weapon in a turn -- so if you were to stand still, you'd spend 10 initiative points, but if you hit someone with a warhammer you spend 12, and if you stab someone with a dagger, you spend 8.  Casting a spell could then be a fixed number based on the spell, e.g. fireball takes 20, flame dart takes 10, and focus takes only 5, rather than the current system which is just a number of turns. 

* Throwing knives seem to not actually do anything once equipped.  I like the item idea though, and it might also be cool if you could give one of your troops (rather than champion) units a javelin item that worked similarly -- a single-shot, short-range missile attack.  Would add some good flavor.

* It would be cool if an archer champion could also have equipped a melee weapon for fighting in melee, since melee damage will I imagine continue being much larger than ranged damage.

* When trying to level up an archer champion, it's really not clear what skills one should be selecting as you level up -- dexterity and the "vital strike" tree seem like they should help, but in practice, having made it to level 15 with 40 dexterity, the only thing that influences how effective your archer is is what bow they have (and I have never found a magic bow).

* I found it too easy to go in and kill AI players by sacking all their cities; it seemed like the bonuses for defending a city didn't provide enough of a defensive advantage. 

* For creatures like shrills that have spells they can cast, the AI seems to have them only cast spells sometimes; often you'll be fighting a group of 3 fire shrills and only one of them casts "blaze" on you.  Is this intentional or a bug?

* It's too easy to murder all the spellcaster-focused champions available for the other faction to recruit.  Since these characters haven't been imbued, they can't cast spells, so they're just a dude with 30 hit points who clubs you with his staff and has good loot.  I think any champion who starts with "Path of the Mage" should start out imbued, and if the other faction tries to kill them in tactical combat before they are recruited, they should be able to defend themselves with spells.

* AI-controlled spellcasters are far too willing to club people with their staffs.  The quest where you go fight a lone "dark sorceror' was really easy, because I had an impulsive mounted champion ride next to him at the start of combat, and then this "dark sorceror" proceeding to club him for 1 damage, rather than doing some dark sorcery.  It was disappointing.

* There should be an option to "auto resolve" a combat without using spells.  Otherwise, you end up needing to fight every battle manually to stop your champions from recklessly draining your mana reserves.  Or maybe it should just always be "no spells used".

* It's too easy to get "first strike" against the AI by placing your troops in a line just out of reach and then pouncing on the enemy when they enter range.  Combined with the lack of retaliation attacks and the high power of certain weapons, the human basically always wins "fair" fights with a large margin.  Similarly, against a weaker opponent you tend to suffer very little damage.

* The bear's "maul" ability should perhaps be capped to 2 strikes, rather than "until the target dodges, even if it takes 15 attacks".

And then I have a few more speculative ideas that I think would help the game:

* I wonder whether one should start out with a bit more in the way of basic units one can build (at least as an option!).  It takes hundreds of turns before it's possible to create _any_ sort of mounted or ranged unit, and that point feels very late in the game -- I generally have a group of 5 level 10-15 champions, etc.  So maybe you should start out with the technology to make "crude bows" or slings or some sort of missile weapon that isn't very good, and then you could have one or more technologies later on that let you build better archers; it would make early game combats able to be more tactically interesting, since at least you could try to "protect the archers".  Similarly, I feel like there should be some monsters who you can't count on to charge you (e.g. because they have a good ranged attack and are relatively week in melee).

* The penalty for having your level 3 city's defenders defeated by monsters of completely destroying the city seemed too harsh -- maybe the effect should be more like killing N people (depending on the monster strength), potentially lowering the city level and destroying a few buildings, before leaving?  That's what it did in Master of Magic, and I thought it provided the right balance of your trying desperately to avoid it but not a "load the game" level failure.  Of course it was always weird that in MoM the monsters disappeared afterwards; maybe a better balance is that the monster hangs out in the city wrecking things at a rate of 1 building + 10 people per turn, and only destroy the city outright once it is out of buildings, but you can rescue what remains of the city by killing the monsters?

* One thing that bothers me about the currently technology progression requirement for monster lairs, is that you build up a big army to kill that drake and its army of pack drakes, and you succeed in killing them, but then it takes absolutely forever before you can actually gain any benefit from having conquered that lair because the lair lets you recruit a really powerful unit, so for balance reasons, you need a really advanced technology to build it. So I wonder whether there could be a lower-level technology that lets you recruit lesser creatures (e.g. individual pack drakes, and then later on a group of 3 of them), but only with more advanced technology are you able to figure out how to keep a full-size drake tame, and train those to join your armies.  It might also make sense to require the drake lair to have been built for a long time (say 50 turns) before you can start building the larger drakes.

* One thing I felt was missing in comparison to Master of Magic, especially when playing as the Kingdoms, is that there are all these cool creatures you run into with interesting abilities, but as the player, you're never going to actually use any of those cool monster abilities, because the only creatures you can summon are elementals, so your units are basically your champions, plus a mob of normal troops with increasingly good arms and armor.  If you play as the Empires, you'll eventually get to train Ogres, which helps a bit, but with the kingdoms, it seems like there should be more variety in terms of cool creatures you can end up controlling, and I'd enjoy being able to control "hell hounds" or whatever.

* This in minor, but army placement on the tactical map could be better.  I'd like it to be the case that when you attack a wildling shaman and two wildling warriors, that the warrios start out 1 tile ahead and to either side of the shaman, so that they're actually protecting the shaman from a rider running in and slicing the leader, like in this diagram below.

            warrior

shaman                                                            ENEMY

            warrior

And it'd be nice if the strongest/toughest fighters in your army generally always ended up in the front row as well.

* In the same vein, I wonder whether the game could allow moving through a friendly creature's square?  I don't think that there's anything that stops a character in the middle of movement, so this should be relatively easy to implement (except maybe for graphics issues?)

* I like how if you let e.g. a darkling lair fester, eventually you'll have whole armies of darklings showing up in the area they control.  Now if only the resulting army sent detachments to seek out plunder ...