Isolating a gene or sets of genes in diseases sometimes isn't enough to determine penetrance - epigenetic factors can have a significant effect. What are the criteria in determining whether epigenetic factors are significant?
First of all, the nature of penetrance is almost entirely unknown. Likely it's a combination of epistasis and gene interactions, induced gene regulatory pathways, developmental noise, and other factors. Epigenetics (imprinting, etc) may have little to do with penetrance, while chromatin structure may be a consequence of other things (most now regard histone modifications, etc, as consequence of transcription rather than heritable regulatory mechanism). Currently, the field of epigenetics is undergoing a (long-overdue) reassessment. Until that happens, anyone who wants to make claims of "epigenetics" is free to, so spurious claims are rampant.