There seems to be an almost unlimited number of ways to create LDPC codes.
There seems to be an almost unlimited number of ways to design LDPC decoders.
There seems to be no way to evaluate how good a LDPC code and decoder are.
There seems to be no way to determine the minimum distance of a LDPC code.
There seems to be no way to determine if one LDPC code and LDPC deocoder is better than another because it is impossible to exhaustively test them against all possible error patterns. One thing that could be done is to evaluate short LDPC codes and decoders to compare one against another and assume that if short ones would behave like long ones.
I've been told that it probably takes a team of from 50-60 people (many with PhD degrees) to design, develop, test and validate a LDPC solution which must cost millions of dollars. Developing a binary BCH or RS solution can be done by one individual. How many companies can afford to develop a LDPC solution? I don't think it would be practical for a small or even medium-sized company.