We have used this technique for MPs! The KBr must be suitable to avoid the H2O artefacts (humidity free) and the sample must be near to powder! The key is that the pellet should be homogenuous (MPs+KBr)!
Personally, I am not a fan of this technique for MPs, because the obtained spectra need a miticuluous post treatment by softwares to have good ones!
At all, I find that the best ones for MPs' identification are RAMAN and u-FTIR.
Adeel Rafiq Flame pyrolysis in GC (perhaps with mass spec) is one good technique. I used flame pyrolysis (with a heated Pt wire and Perkin-Elmer GC's in the early 1970's) and a whole range of polymers could be identified including mixtures.