This is exactly what I was trying to point out. Steam's developer services can save a massive amount of time for developers, time which is especially valuable to indie studios. I still feel that the 30% is steep, I'd prefer if Steam took a cut based on how many of their underlying services you used, but its wrongheaded to deny that Steam provides many useful features for developers that can save a lot of time.

https://partner.steamgames.com/doc/features

I think such model can lead to messy scenarios. Say you start without cloud saves and sell million copies. Then you add cloud saves. Now should your commission increase on past sales meaning that for while you make nothing? Or should it only apply moving forward and on future copies sold?

And I am absolutely certain that some developers would exploit this in someway.