The issue here is that those bugs should never have made it to 'prod'. For absolute die hard fans who say "Gnome or bust", you encounter a bug, you report it, track it, get it fixed, and things are better, until next time the same high level symptom - UI slowness - is introduced, through a different mechanism making it a "different" bug.
Thing is, not everyone is a die hard fan. Developers on Gnome, by definition, are, but outside of that, customers are going to go with what works. And that's not Gnome. Customer's don't care that the UI slowness in version 1 was caused by X, and UI slowness in version 2 was caused by Y. There was UI slowness, which made it unusable, and, well, now they're using XFCE or OS X or Windows. It doesn't matter that there was a bug and it was fixed a year ago. The problem is that there was a bug a year ago, and that particular customer is gone and can't be recaptured - they've got a solution that works and they're not looking to switch to a different one. Especially one that has a history of problems.
Thing is, not everyone is a die hard fan. Developers on Gnome, by definition, are, but outside of that, customers are going to go with what works. And that's not Gnome. Customer's don't care that the UI slowness in version 1 was caused by X, and UI slowness in version 2 was caused by Y. There was UI slowness, which made it unusable, and, well, now they're using XFCE or OS X or Windows. It doesn't matter that there was a bug and it was fixed a year ago. The problem is that there was a bug a year ago, and that particular customer is gone and can't be recaptured - they've got a solution that works and they're not looking to switch to a different one. Especially one that has a history of problems.