According to this report, there is inconsistency between DAS,MCM,XSDB, twikis and GenXSecAnalyzer results.
They are working on some tools.
https://indico.cern.ch/event/916121/contributions/3900342/attachments/2052768/3441201/2020_06_08-XSDB-GeneratorMeeting.pdf

Something we might want to consider for the future, aka RunIII, is to build such kind of database.
This is a nice way to handle/store/verify these values. ;) Maybe we can take inspiration from it.
https://github.com/TreeMaker/TreeMaker/blob/c95b539b050f05789e54d6ec5ec8a61132d75077/WeightProducer/python/MCSampleValues.py