complete docs pertaining to vcm

[alecristia]

in particular, how vcm integrates into a pipeline (does it rely on yunitator output? can it take a different input, such as textgrid or rttm with a "chi" speaker tagged?)

and how it can be evaluated (using evalDiar or not?)

[riebling]

I just learned yesterday, there is only 1 line of RTTM produced, because the result applies to an entire audio file. "aha!" :)