A free replacement for Flash is likely to become less useful in this situation, as other non-proprietary solutions are becoming prevalent.
A "unique to free software" solution makes it so people want and need to switch.
Between the two, I think creating incentives is more powerful than removing obstacles. (Given strong enough motivation, people will overcome obstacles. Lacking strong motivation, why bother?)
Because their goals include that everything possible should be possible for to do with free software. It is by definition a game of catch-up.
Let's innovate people...
The goal of the FSF is to promote the use of free software. If people won't or can't use free software because they need some proprietary program, then the goal of the FSF becomes to eliminate that need.
Sometimes this involves writing software that is better than proprietary equivalents (c.f. emacs and gcc in their early days). Sometimes it means just cloning what's there so you have something to use (most of coreutils falls into this category).
Why is having a free program that solves a problem that's already solved more important than creating programs that solve new problems, but that also happen to be free?
I find it hard to believe that the most important free software project in the world is a player for passively consuming content in a proprietary format (Flash), that is A) dying and B) more for eye candy than anything useful.