Move thread execution detection into create loop (as we can't restart dead threads)

Move thread execution detection into create loop (as we can't restart dead threads)
Altered create loop to run all the time (instead of only at station creation)
Fixed issue where streams would not be closed if an error occurred in playlist advancement
Fixed issue where saved playlist position could result in an index overflow
Added maxretry value (determines how many times to attempt a station thread restart)
Fixed minor documentation things

Signed-off-by: achbed <github@achbed.org>