Is the current situation that your code from post #39 works against your test transmitter code which appears in post #35, however, it does not work against the transmitter which your professor has provided for this exercise ?
If that is the case, it is very possible that your professor has done a few tricks, within the limits of Manchester encoding, to make this exercise more challenging, such as varying the timings within a certain tolerance etc. which would require some better attempt at clock recovery/re-syncing in your code.
Anyway, can you somehow get an oscilloscope picture of the output of this "difficult" transmitter?