Right now we're doubling up on a lot of hashing algorithms.
The WebAssembly implementation is a good first pass but if we're keeping single algorithm imports (e.g std/hash/sha1.ts) then we should probably generate WebAssembly backed classes for these as-well.