I'm on my 3rd game now and I have a couple of observations about the new starting units concept. Twice I've bought a hero on turn 1 and once I didn't and used the extra gold for better production.
Things I like:
I like the idea of getting multiple units (3 or 4) to start.
I like the idea of getting initial gold to buy a hero/production as the player wishes.
Things I don't like:
I don't like the idea that it's really biased towards spending 500 on the hero (because you get an extra unit - the hero and because Piranha raised the minimum gold cost for a hero from 200 to 350).
I don't like the idea that the amount of gold left to spend virtually always is used to buy the best available unit (Giant in a medium income and Pegasi in a high income).
So instead what I'd propose is that you:
1) Set the initial number of armies for all players to 1-2 instead of 3-4.
2) Allow the player to buy the hero for 500.
3) Allow the player to buy production. With the difference being that each production unit you buy immediately adds 1 unit of that type to your initial starting armies.
Now you can potentially buy 4 extra units on turn 1 if instead of spending your gold on the best unit (Giant or Pegasi) you buy a bunch of low/average units . You also are now not penalized for not buying a hero on turn 1 because if you instead buy a Pegasi (as I did) you'd get a Pegasi on turn 1 instead of having to wait till turn 4 to get your first Pegasi.
I think this will produce a much bigger variation of initial spending of the gold. Because right now it's pretty much going to be Hero + Giant (Hv Calv) production on medium incomes and Hero + Pegasi production on high incomes.
KGB