But such specificity is not nearly as problematic as the unnecessary differentiation between speech-to-text and speech-recognition, right? Also, you say "few applications (…) offer voice commands but no speech recognition"; OS X only since Mountain Lion includes dictation, and even that is performed on servers, not locally. So it's distinct stuff. Dunno, it doesn't seem useful to merge tags like these where distinction does exist. It's confusing to new or occasional visitors of the community (like me), and does little to help regular users.

Speech recognition includes voice commands, but is not synonymous. For example, dictation is speech recognition, but is commonly not considered the same as voice commands. Right?