it's so bad that I can probably count on one hand papers I've read using LCGA or LCGM models that actually use them in a reasonable way (e.g. "we think they're a handy way to describe variation in change over time"). Almost every paper ever mistakes the classes for "carving nature at its joints"
