Probably because these things normally tend not to work :-) so people don't care (maybe). I don't, at least.

Trainable Voice Reco works great, in general. It's much easier for a computer to examine two sound files for similarities than it is to break up the sound file into utterances(syllables) and try to figure out exactly, word for word, what the user said. I wouldn't be surprised if the MK2 Voice Reco worked perfectly for its user within a few days.
So can the microphone be retrofitted, too?

_~= Dearing =~_
"WAY too happy about having #99."