Observation #1 --
When I run:
http://developer.echonest.com/api/v4/song/identify?api_key=
When I run without any version parameter, I get the full identification response in JSON.
Observation #2 --
I built and ran the iOS code from rexstjohn/echoprint-ios-sample on iOS7. Here, when I pass the statically-defined PCM code to the same server, I find the same behavior as noted above - i.e., I must run without version number to succeed in identification.
However, most importantly, as many others have reported, I still cannot get any song to ID via the microphone capture. I have been trying many songs played through iTunes over the air, after verifying that the song was in your list of songs from the DB.
Help!