While the board concept is really cool, I think the starting defenses might need to be upped a bit.  I've found that it's possible to eliminate someone before they even have a chance to take a turn.  I haven't finished any of my games yet, but I'll post one in a couple days where I managed to take a neutral capital on my first turn (the neutral could have just as easily been a player that hadn't taken a turn yet).

It is possible, but highly improbable, statistically speaking.

If the attempt works then you'll definitely have an upper-hand, but if it fails then you've just shot yourself in the foot.

There is always gonna be that possibility in games with capitals but there is a max of 10 in the invaders and those 10 have to go through 30 initial units in the shields at minimum and then whatever is in the cannon itself. With return to attack and 2 fortifies you could put 20 into the 1st shield spot you take and go from there but I'm not sure how you get that many units in one spot.
OK, if you are playing a 2 player game on the map it's anybody's guess that will happen. It was meant as a 5 player map and it says that in the description. I opened it up for other amounts of players because the fog makes you waste armies on the neutrals perhaps. It wasn't calculated around anything but 5 players (Yes, I do think before I make a map although it's hard to tell sometimes).
Anyways, if it is a less than 5 player game you are getting more invaders at the beginning and that is gonna mess with the balance of the units in the shields so you are right, there is a flaw that way.
You all decide, should I keep it as is, flaws and all (much like me) or have it switched to a 5 player only map?

I'm going to go ahead and say this map would be better if you turned off assimilation as well. When someone leaves a shield open it hands the game to whoever attacks that person - assimilation is very powerful.

I would rather see the map default to 5 players, but the option to play with however many you want.

Toaster, I don't think the first-turn elimination scenario is as improbable as you make it out to be.  I've now seen it happen in 2/4 of the 5 player games I've played in.  I admit my sample size is small so far, but I think the odds are pretty close to even.

This is pretty close to the optimal opening move to eliminate someone:

1. Place 7 units on an invader that's in the 2nd row from the bottom, above an enemy's cannon.  Place 7 units on one of your shields.
2. Take out the 3 neutrals below your invader with your cannon.  This is 10v3 and shouldn't be hard.
3. Move your 10 invader units down to above the enemy's shield and attack.
4. If you get slightly better than average luck, you can take the first shield territory with those 10 units.
5. Transfer all 10 units from one of your shields to the shield territory you control.
6. Keep breaking through shields, and when necessary transfer 17 units from the shield territory you placed your 7 units.

This ends up being 10+10+17 = 37 attackers vs. 40 defenders, so it's definitely worth the risk in my opinion.  If you succeed, you now have twice the unit production of other players and are much harder to eliminate in a single turn (given the attack limit).  This sets you up to eliminate another player every turn until you win.  If you fail, you can still hope that you don't get targeted by another player for elimination (or they don't take the open path through your shields), and you have a good chance of taking out the other player on your next turn.

it'd be an entirely different game if there wasn't return from transfer...

I was going to crunch numbers in a retort- thanks to Norseman for supplementing my motivation. Agreed on all counts.