> What gives? Is this just part of the growing "corporatization" of Linux
I would argue that this is the case, and if it's not the case then mostly because it's not becoming more corporized but because it always was.
I would argue that there are three popular reasons for open source development, that there is a huge difference in user impact between these, that corporate interests have been the main driver behind open source, that having ones own preferred solution allowed to take profit from the competition, and now that the profits are dwindling distinction between competitors becomes less important than saving costs, and therefore everybody focusses on whatever appears to be as the winner in the current topic area.
Reason 1 for open source development, and probably the original reason, is art. People wanted to scratch an itch, they themself felt. They don't care much about gaining any additional advantage so they share the whole thing and not just the binaries, maybe by doing so gaining some recognition by other smart people. These artist developers usually hate the normal user and despise the disrespect these bring towards their piece of art. They often don't care about the trivial problems the user has, like the need for a complex yet intuitive UI. So users will often refrain from using these tools. Making them mostly unimportant at large, if not someone else comes along and uses the knowledge gained from it for another category of open source.
Reason 2 for open source development is the goal of social and political influence, which drives a person or group to develop tools to fascilitate communication between supporters and fascilitate spread of the culture. One example I see here is the scuttlebut network, which if you analyse closely, is not really good for everybody but mostly used by a group of New Zealend developers to spread their ideas of off-grid lifesstyle and something they call "solar punk". This is by no means criticism. It's a great tool/protocol and the ideas they spread with it also don't seem harmful to me. But without the desire to influence the political landscape they probably wouldn't invest that amount of time and energy to build their social network. This kind of project can gain quite a lot of influence on users at large. For instance all these 2chan,4chan,8chan imageboards might rival professional social networks in size and usability. But usually it's limited to people who support the political goals or at least can accept them. For instance it's unlikely that a group of feminist game developers would choose 8chan as their platform after it became a platform for the anti-feminist gamergaters.
Last but not least stand the group of open source project which directly or indirectly are for profit. For instance Linus Torvalds is one of the best earning people on this planet. Multiple companies have formed around linux distributions. And most of the commits nowadays come from people who are paid to do them. As said as it may sound, this is the most influential group. It's simply a fact that user-focussed development can be a daunting task and most people wouldn't do it if nobody paid them for it. So the development is also built around market dynamics. When cool UIs gain user attention most distros will focus their resources on one desktop manager that separates them from the rest, so that users who like these distinctive features will come to them. And when profits shrink they will look for whatever desktop manager is the current winner and jump on that band wagon, so they can reduce their development investments and focus these on other topics.
If we can agree on these three categories I think it's clear that the last one is the one that gains the most users, and therefore will also influence open source in general the most. Also it might have never been different from that, and as long as the getting is good it's also not to the users disadvantage.