I'm having a lot of trouble figuring out how to calculate future total mining using historical difficulty in my excel sheet. I assumed that if difficulty increases 10%, I would mine 10% less... But this does not seem to be correct
Why does it seem incorrect to you? the rewards will follow the difficulty (not 100% but that's a different topic), so if you assume a 10% increase in difficulty every month, then whatever your Excel sheet showed is correct.
The question is, why would you use 10% per month? the past 12 months had a total difficulty increase of 53.6%, that's more like 4.5% a month than 10%, also this can't be linear progress, just because it went from 1T to 50T from 2017 to 2023 which is an increase of 4900%, it doesn't mean that it will go from 50T to 2500T by 2029, that is not physically possible, not even remotely possible.
Now doing all these projections and estimations for what the difficulty might be in the future is pretty much useless, it's all directly connected to the price of Bitcoin which no code or equation on earth can predict, logically, it's now harder to make a 53% increase in difficulty than it was last year, cause a 50% from 100EH is only 50EH, but a 50% of 300EH is 150EH which is 3x more.
BUT (with all capital letters) it could still happen if
BTC goes to 100k in the next few months, you can almost guarantee that a difficulty increase of 4900% in the next 6 years is impossible, but you can't be certain that a 100% increase next year is not possible, as far as mining is concerned, what happens next year is much more important than what happens 6 years later, the short time frames are what matters, and predicting those with a good accurate is impossible, whatever gear you buy today will probably be dead or useless 3-4 years from now, so all the math you do today in respect to 10 years later is a waste of brain cells and computer resources.
But anyway, the way you do the math is correct, maybe not 100% accurate but correct.