The problem is the arbitrary way the funding is allocated, find something interesting to companies du jour with a clear path to funding and you're golden.

If whatever OSS is too obscure to be noticed by non-techies but still fundamental (think OpenSSL, libxz,etc) it's more likely to lead to burnout far before anyone wants to put in any sane money (curl is one of few counter-examples but that hasn't had a straight journey).