Free stuff is OK but relies on someone implementing some feature or fix because they feel like it. A commercial enterprise may not feel like it either, but at least you can actually withhold money (or not spend it) to let them know how you feel about it. Also, currently, the free stuff isn't up to the paid stuff - Kicad, for instance, won't run on Windows 7 and the authors don't give a toss.
This relationship goes both ways. Users are generally not an asset to open-source projects, so contributors couldn't care less how many people are using it. If anything, users are a liability; they generate support requests and asinine complaints like 'it doesn't run on an out-of-support, 13-year-old OS that, whether we like it or not, is no longer supported by one of our core underlying tools'. So yeah, the people that actually are contributing to the project aren't going to put their effort towards features targeted at making users that can't be bothered to even maintain an up to date system happy. Such features are of little to no benefit to the contributors, who have no problem running a modern Python on their machines (probably on Linux...). It might even have concrete drawbacks to them, like being stuck on an ancient Python version. Who wants more whingey curmudgeons that refuse to get with the times associated with their project anyway?
On the other hand, individual customers, especially hobby users, have no weight with commercial entities either. If you want something different like say a perpetual not-cloudy license of their software, or support for Linux. Or anything, really, your option is going to be 'pound sand'. At least with open-source offerings you could potentially implement such support yourself, find an interested developer, or pay someone to implement it for you. It even has a good likelihood of getting upstreamed.