If this is a potential problem once GPL-2.0 is changed to GPL-2.0-only, then it is currently a problem.
Yes indeed, that's my point :-).
And perhaps by altering the current identifier (GPL-2.0) to be more explicit (GPL-2.0-only) we will expose just how often GPL-2.0 has been used incorrectly.
The tools are currently *required* to be incorrect, because they cannot report the information they have ("I have GPL-2.0, and I don't know if 'or later' applies"). Neither the proposed "GPL-2.0-only" nor "GPL-2.0+" correctly represents the information they have. Tools will have to output *something*, and whatever they produce will dilute in *practice* the strict meanings of those license identifiers.
