Delete (and map) entries that have equivalents in other datasets

[strogonoff]

• Map NIST/FIPS and remove from this repo #8
• Map W3C standards and remove from this repo #7
• IEEE (see example below)
• Others?

Case in point: IEEE 802.1QBV_2015.

• The relaton-data-ieee source seems to contain a better version of what seems to be the same standard:
  https://dev.bibxml.org/get-one/by-docid/?docid=IEEE+802-1Qbv.2015&doctype=IEEE
• For comparison, this is the version from bibxml-misc: https://dev.bibxml.org/get-one/by-docid/?doctype=IEEE&docid=IEEE%20802.1QBV_2015. Note that it specifies an apparently nonexistent or malformed DOI, and offers less complete metadata overall.

Proposed plan

• Strip from this dataset all documents that have more complete equivalents in other datasets. (I don’t suggest to delete them from any underlying XML source; just make sure they don’t get into relaton-data-misc. I don’t know the best way of doing this, perhaps by filtering out certain filenames in the GHA workflow?) This would mean 1) old xml2rfc paths still work (via bibxml-data-archive fallback) and 2) GUI shows a newer version. I suspect there are no downsides.

• Iff removed document corresponds to a legacy path (that is, a path that exists in bibxml-data-archive), we can additionally map that legacy path so that they don’t fallback and return more complete data.

Asking for comments @ronaldtse and @andrew2net.