How many apps are successful enough to meet the revenue thresholds (over $1mil), but are sub $1 ARPU, and aren't ad revenue driven? I'm sure there are some, but that's going to be a small number, and there likely can be some carve out for them.
Those thresholds are just the point when you'll be net negative. There's still a large range above that ARPU where you're making less money than before or with a different engine. Cult of The Lamb has already announced they will delist their game, and the Slay The Spire dev team have announced they're halting development on their new game to port it to a new engine.
Those are exactly the types of games that would be much worse off under an Unreal style royalty.
Cult of the Lamb is $24.99 on Steam
Unreal royalty is 5%, so $1.25 per sale
Unity is $0.20
So royalty is ~6x more
If there was a reasonable way to count installs (I don't think it is technically possible), install fee is much better than royalty for most devs who are not F2P apps.
This is not an accurate calculation as it assumes a single install per purchase.
Let's say I really like this game and install it on my laptop, desktop, and steamdeck (3 installs). Now, let's say I upgrade from win10 to win11 (4 cummulative installs). Alright, now I'm bored of the game an uninstall it. A year later the game recieved a new free DLC including new content. So, I install on my laptop, desktop, and steamdeck again, after I have updated both PCs from win10 to win11 or made some other hardware change (7 cummulative installs). And repeat this process every 2 years.
Unity's price is now much more than Unreals as I'm now costing the dev $1.40 after just the first update. Every subsequent update means I'm costing the dev another $0.60.
This may seem ridiculous, but anecdotally, it seems relatively common.
Again, the issue is not with Unity wanting to make more money, but with how they are trying to achieve this. Install based cost means you cannot estimate tooling cost as it is impossible to estimate situations like above.
