Enlis flagged some SNPs in my 23andme data that I think are miscalls (i.e. out of the almost million SNPs, Enlis zeroed in a couple that were probably mislabeled by 23andme). The Dante Labs data should be more accurate, so you probably won't have that problem.
Also I'm not sure how often the Enlis database is updated with new research, or what their threshold for reporting is. Promethease has flagged things that Enlis ignored. Overall I have the impression that Enlis is more clinically stringent, mostly noting things that are well-documented; and Promethease casts a wider net, showing newer research and smaller studies.