Again, with what you laid out, i assure you, go with axing the bonus to mines on gunpowder. Much less damage done.
Now at least i can understand where the idea came from.
But you obviously didn't pay much attention to the flow of a real game. + Asking the community prior to the change (like you did with magic and to a far smaller extent in importance chariots) would have been vastly more helpful for even your own resources.
Mines with +2 Hammers are not! far to strong.
There is a unique drawback, food shortage/having to compensate with a farm / slower growth which is downright impossible to mitigate early-game. Plus there is a substitute which is exist, earlier and for free. Forests. (for elves it isn't a substitute. But there balance is possible with far less collateral damage.)
Add into that unlike the other improvements its strongly dependent on terrain (hills only. cottages go everywhere there can be food) and in forrested terrain you also need a tech to be able to improve in the first place
(not so huge an issue since mining is also the tech for chopping forests. But jungle hills are still exist. So its still the "latest" of the early improvements without anything near / in sight to improve them unlike the other improvements, especially farms. And the next one who'll propose to nerf farms ill metaphorically lynch via the net.
)
Now they even become more limited to just unforested hills without nonmine resources? Looks like it...
Also Mines are the only possible and thus viable improvements for plains-hills, Other things just don't work (windmills excluded but they seem like they will become the new mines, with mines now dieing a horrible demise...).
The existence of forests (which give +1 hammer and! a bit of health without any of the worker turns) shouldn't be mandatory and clearly superior for production.
Add into that the added usefullness / value of forests in FFH2 beyond that of vanilla (i doubt need explanation.) and that chopping is inferior and you are now in the odd situation that forests will be overpowered...
Please don't just go with a chart / maths without minding the wider implications. (on grassland-hills mines often lose out to even cottages with +2. Let alone other improvements.)
Also as is grassland is already far! superior to plains.
Now after such a change expanses of Plains are a downright death-sentence even for production-centers. (Aren't grasslands usually more forested as well?)
Read: As useful in overall game-flow as tundra has been before.
How odd is that? And totally counter-intuitive (as in makes the mod harder to play / learn, for what gain in terms of gameplay exactly?)
Yes I'm rather sure of / understand what that means and i mean it. Take a few seconds to contemplate the implications. (On a related note Forested Tundra will be about as valuable as un-forested Plains which makes the comparison even more fitting. Since its quite realistic honestly. As in often seen in many real games.)
Yea turning half the world into the equivalent of tundra sure! will make the game more fun.
You'll not help the balance of the game if you'll inflate the randomness of the start even further. And make grassland starts the only viable option with anyone else dieing a painful demise. Poor AI who can't relocate their settlers and doesn't mind such things on expansion.
So all in all as a pie chart or bar-graph for balancing your concept might look! nice. But in terms of a real game its abysmal and about as useful as i outlined in the first post. (and i'll leave out any notion of uncertainty this time now seeing where the idea came from.
The Numbers / math looks! ok / understandable but! the conclusion / chosen way of solution is just! plain! wrong.
2+1 = 3 is to strong (? and a very strongly bolded "?") so we'll make it 1+1 = 2 instead of 2+0 = 2?
Go with 2+0 = 2. Trust me. If you really mind the wider implications and you'll see why very fast.)
On a bit lighter note: Take heed to the insights of the global financial crisis. Statistical maths / pie charts / bar graphs and real life just don't mix...
And to reiterate the problem with the improvements was not! that the early game improvements were to strong (By what account in the first place? Besides purely! in comparison to the other improvements which can actually be moved up to make them a comparable option. That sounds so much like balancing for balancings sake without a real need for the way you proposed in a real game / in terms of gameplay.) and not! that the early game seemed not allright.
The problem was that in late! game the late game improvements have not been viable in comparison to the then-improved early game improvements.
I can see your point that an improvement which is available in early game and stays powerful because it still improves in power later. But then the early game is precisely not the only or best place to look for balance even if! you to stick with nerfing the early-game option. If a figertip has a scratch don't amputate an otherwise health arm.
The balancing of the game / flow of the game was not! off or broken by any means (i seldomly heard that there is a problem if you purely take farms / mines / cottages by themselves).
It was the inter-balance of the different improvements especially in the mid-late part of the game. And it was a mediocre issue not a huge one (in terms of real gameplay. The game didn't feel broke because of it. Its just odd mechanically.).
?
But surely don't even need the slightest boost. Upper end of what should be possible) So:
). If implemented i can only repeat myself with the water-techs like fishing / sailing / optics being a good place to put it. 
/+GPP) vs. plains mine (+1
/+2
