This seems fine to me. I suppose you might benefit from automatic build dependency generation that's supported by recent versions of RPM, but certainly that's an entirely different issue.

I think this is fine (I know little about R), but I'm not sure about this bit:

+Typically, R extensions do not contain license files per CRAN policy.
+R allows a set of open source licenses and R extensions just declare which one they adhere to.
+Following this policy, we do not require upstream R extensions to add license files.

Yes, so it refers to this piece of the Writing R Extensions manual:

The mandatory ‘License’ field in the DESCRIPTION file should specify the license of the package in a standardized form. Alternatives are indicated via vertical bars. Individual specifications must be one of

• One of the “standard” short specifications as made available via https://www.R-project.org/Licenses/ and contained in subdirectory share/licenses of the R source or home directory.

So TL;DR, R itself provides a set of licenses, extensions must pick one. When you have an extension (R package) installed, R itself is installed, so the license is present.

Please note that this has been done in this way forever, and going against this upstream policy, apart from redundant, would be a pain, because almost no R package includes any license file. I'm not changing anything, I'm just clarifying this aspect.

Was hoping to get to discuss this at the meeting, but we ended up with a long discussion about the joys of rpm parsing.

I did let everyone know about this PR though, and I said I'd merge it on Friday unless someone spoke up about it. So countdown is on.

Thanks.