Raymond:

> Functions as simple as shuffle() eat through the possibilities very quickly.

Can you elaborate on this?  Are there example scenarios where seeding with 32 bytes isn't likely to be enough?

In the case of shuffle, for a large list, if you do a seed followed by a shuffle, the restriction to 32 bytes is restricting us to 'only' about 10**77 possible different shuffle results.  It's hard to imagine a situation where that would be a problem.