One other option to look at, if you aren't already, is to freeze new feature development and allocate most resources to making the platform solid. You didn't state that in your list above.
You may be correct that you've done a lot to stabilize it during the last year, but how does that compare with the amount of new features you've added over the last year? Maybe Facebook just isn't good at producing bug-free code, and it's time to improve things you do before actually shipping a feature? Continuing to "ship fast, break fast" clearly hasn't worked when it comes to Platform - are you doing anything to fix that?
As an anecdote, Microsoft had a similar problems with Security of their OS - it was a much more severe problem. They took a similar approach to the one you've outlined, for many years - furiously developing new features while scrambling to patch security. However, in the early 2000s (2002?), after a particular severe worm hit computers all over the world, they completely froze all new development on the OS. They did their famous 'security push' where almost every engineer focused on security for many months. They revamped tools and invented technologies to make sure products are more secure out the gate, and the result was a resounding success.
Developer frustration with the Facebook Platform is very real - a competitor with a stable API might actually receive a lot of support from the community. Maybe Facebook needs its own version of Microsoft's 'security push'.
You may be correct that you've done a lot to stabilize it during the last year, but how does that compare with the amount of new features you've added over the last year? Maybe Facebook just isn't good at producing bug-free code, and it's time to improve things you do before actually shipping a feature? Continuing to "ship fast, break fast" clearly hasn't worked when it comes to Platform - are you doing anything to fix that?
As an anecdote, Microsoft had a similar problems with Security of their OS - it was a much more severe problem. They took a similar approach to the one you've outlined, for many years - furiously developing new features while scrambling to patch security. However, in the early 2000s (2002?), after a particular severe worm hit computers all over the world, they completely froze all new development on the OS. They did their famous 'security push' where almost every engineer focused on security for many months. They revamped tools and invented technologies to make sure products are more secure out the gate, and the result was a resounding success.
Developer frustration with the Facebook Platform is very real - a competitor with a stable API might actually receive a lot of support from the community. Maybe Facebook needs its own version of Microsoft's 'security push'.