I'm going to echo many others here, I fear...
Play for any extended period of time and the game will crash with an OOM error.
The AI remains atrocious, the tactical worse than the strategic. There's nothing like attacking Lady Porcupine the mighty channeler some 300 turns into the game and take on her little empire, only to see her charge into melee and get killed by the retaliation attack when she attacks a group of soldiers instead of casting spells. And do do it three times in a row as her cities are taken one by one. Somehow or another, despite controlling multiple shrines and way more than enough time, she either hasn't learned any spells, has spent all her mana on the strategic level (doing WHAT? God and presumably the Stardock developers knows - perhaps she's casting Alchemy or any of the other worthless spells that exist as traps to drain the unwary of mana?), or the tactical AI just decides against casting spells.
Evening the odds by not using any magic oneself and making the game essentially "Elemental: War Without Magic", the AI remains incapable of putting up a fight against a warfare/civics focus for long games or against exploration[champions]/warfare[champion equipment] focus for (moderately) short games.
Something I don't know if others have mentioned (or experienced) is a bug that has been in the game since release and expected to be gone by now - occasionally the wrong city name and stats are displayed on the strategic map. I.e. I might have a city, MyNiceCityWithAGoodName, and when panning the strategic map to its location it will display as AnAICityWithAStupidName and that city's stats in the colour of AnAI. Selecting the city will refresh the map with the correct name in my own colour and with the correct city's stats, so it is purely visual - it is just that it looks so darn stupid to see my cities occasionally mislabeled, particularly when it is one of the linguistically silly AI city names.
Example (Newly conqurered Raracar [west] stats painted on two cities):
I lashed out against the spells in 1.2 and I did so for a reason; Ignoring for the sake of argument how bad the AI is at using them, there are severe balance issues. Take a good hard look at the city enchantment spells and ask yourself "which ones would anybody even consider using? Which one is worth both the casting cost and the upkeep cost?". At what point in time is it correct to spend the arcane research on advancing to spell level 2, learning Invisible Workers, and then spend 16 mana and 1/turn on getting +1 material/turn? There's a certain point where the opportunity cost of casting a suboptimal spell transforms it from being an interesting option that may fit into a specialized strategy or be effective in a limited number of situations and into a trap for the unwary, preventing the player from casting any the good spells should he waste his mana on it.
The same goes for the other categories of spells, of course - I chose Alchemy as one of the most egregious examples at start - 25 mana for 25 gilders. Is it even possible to have both researched this spell and be in a situation in-game where it makes sense to transform mana to guilders one to one? Given how hard mana is to come by compared to money it seems unlikely.