New hashing functions - most notably Whirlpool, and the remainder of the SHA functions - have been added, along with RIPEMD.
Documentation is lacking, but efforts to fix this are underway.
There are only a few strong hashing functions not yet implemented in MHash. The addition of at least some of those, plus the docs and some code auditing will bring the version to 1.0.0. Examples include Cellhash, Parallel FFT-Hash and Subhash.
The main efforts from 1.0.0 will be on speedups and security auditing. The idea here is that we want MHash to be not only the most comprehensive hashing library, but the most trustworthy as well. New algorithms'll still be added, if anything new or exciting comes along, provided there are grounds for saying it's useful, fast and secure.
Faster implementations of existing algorithms are ALWAYS welcome.