I did an LCGA (Mplus) with three time points in the last couple of years and would enjoy any more recent papers, but I am also curious if anyone has done LCGA in R and what they used.
I've never done a LCGA in R, and OpenMx is the only package that comes to mind off hand since I've seen an example of growth mixture models (see http://openmx.psyc.virginia.edu/docs/OpenMx/latest/GrowthMixtureModel_Path.html for a GMM example), but I'm not sure how Open Mx compares with Mplus.
Last I checked I couldn't find any good sem packages that handle categorical latent variables very well between lavaan, sem, and OpenMx. If anyone knows differently I'd be interested to know.