Most heavy tasks are not implemented in python so even a 4x speedup in an implementation doesn't matter for most wrapper code plus pypy isn't fully compatible with cpython.

And fewer and fewer people are building large business logic applications in Python (Django is dying out, for instance).

I think pypy is still probably the second most common python implementation though.

I suppose - though we still seem to have interest in a 'faster' python, and its more compatible then a lot of the alternatives. They had just gotten SQLAlchemy(?) working with it when I moved on to other stuff, so I never got a chance to try it out for 'real' work...

I have to admit though, I don't reach for python as much as I used to, so I can see your point.