I'd say it's expressly intended to be limiting

My interpretation of your concern is that you fear that there will only be one valid starting strategy - build as many cities as you can. I think Stardock's intention is for their to be multiple starting strategies that work and so spending all of your essence to build cities is only one path.

I wouldn't be surprised to find that Essence costs are one of the differentiating characteristics of the faction that you choose. So maybe civ X has a cost of 6 Essence/city but makes up for that with the ability to invest cities with Essence that improves their productivity, or they can make more powerful heroes, etc.

Yep, I am talking about limiting the starting options. I still trust the frog, so they probably have a plan.

It just seem so strange that essence is needed for starting cities, but essence will be very limited in the start.

We do not yet know what the other things are, besides starting a city, for early game mana uses.  We may find that a good strategy is to start one city and then use essence to research spells, or summon a troll family, or create a magic stave, or to do something else.

Until we have options to compare we cannot really assess.  On the other hand, I do like the limitation this imposes on the classic strategy of starting as many cities as quickly as you can, that gets tedious.

I expect to see many more uses for Essence as the beta progresses.  If nothing else I believe it's been stated that the sov's remaining essence is the limit to how much mana the sov can use in spells, so if you want to cast uber-spells later in the game you'll have to be careful with how you use it.  Also, as valhur noted, it doesn't cost any essence to create a city on reclaimed land (land within your area of influence), so it's quite possible to start with a single city and expand slowly.  I expect that this organic growth will be where the vast majority of cities are founded and we will only be spending essence to claim vital resources that we expect to be contested by an enemy or are quite far from our core territory. In the long game I expect spending essence on cities will probably be a pretty bad choice.

I'm expecting essence use to be one of the core strategic choices we make and as we get more options for using essence I plan on trying out strategies using each option exclusively (including not spending any and just going spell-crazy).  I'm guessing that making magic items may require essence and since those were among my favorites in MoM I'm guessing I'll be the type to do lots of that eventually.

I'm hoping that we'll soon be having a discussion on what we'll be able to do with Essence, how much we will get and how to keep it balanced.

Well, this is what I see happening. It cost Essence to build a city. Say you build a city that is in a good position but some improvements you want are still a few or more then a few squares/hexes away. Now you have to build another city pretty close to a city you just built so you can get those other improvements.

How will this affect Game Play?

We're going to have too many Cities sprouting up and our Sovereign's will be Essence starved from building so many cities to get improvements.

How can we fix this? I can think of two options so far.

Reduce the cost of Essence it takes to build a City...Or....Make it so we can build a improvement on land that has been reclaimed by our Sovereigns "Without" the need to build a city next to it. So long as that Mine you want, or that Shard you want, is on land that has been reclaimed that spread from a city you built, you should be able to claim it, build the improvement on it, then run a road from it to the near-by City.

Honestly though I think it's going to need some play testing before the Devs will realize this and make it so. Allowing this will also reduce city sprawl.

Agreed. I was going to mention that in my post that it will need proper balancing after all the systems are added. I didn't because I figure that's still months away and it will be something that has to be done when it's all completed.