I just read "Goodbye, OpenStack" (thanks for the link @David Thompson) and it has me thinking about a number of things:

Those I know who work on OpenStack complain about its insane complexity; there's even a comic about it. OpenStack seems like a lot of things at this point. A good number of them seem like things I want to succeed: an API for on-the-fly-provisioning seems important, with a free software backing. That seems great. So does the storage system Swift... we even have support for that in MediaGoblin.
I've been long jealous of projects with tons of funding. MediaGoblin and most of the projects I actually care about are desperately resource-constrained. It's hard not to look at the sea of VC money out there and not feel like I'm hurting the project by not trying to tap into it, but then I think of all the compromises I'd have to make, and how much doing such things seems to affect real communities. But wouldn't it be worth it to be able to pay more people to make the dream come true? But looking above, OpenStack is pretty much the dream machine of money in the free software world. Having a ton of money pouring into it doesn't seem to help its coherence maybe?
OpenStack's decision to use permissive licensing is frequently lauded as one of the reasons it's so well funded. That post seems to indicate that pretty much everywhere is running a proprietary fork, which makes OpenStack sound eerily similar to Android.

Maybe there's some value left in my stubborn, community-over-money, pro-copyleft direction. Then again, being a resource strapped outlier isn't always much better than the above. (And I'm still jealous of all that project money.)