Identify concepts where we use MeSH as preferred identifiers and see if we can replace them with better ones

[gaurav]

As reported in https://github.com/NCATSTranslator/Feedback/issues/355, MeSH is not as good a resource to point to as some other identifier repositories. MeSH also tends to score low on the list of preferred CURIE prefixes, so a preferred MeSH identifier is probably a sign of a bug (e.g. #215). The Babel Validator used to generate prefix counts per compendium file, and of course that information is also available in the semantic_count database on NodeNorm but isn't publicly exposed. So we will either need to build TranslatorSRI/babel-validation#28 or expose the semantic count information.