If you or
@Evan Atkinson use eBird or a similar system, I would be very tempted to use the slash Yellowhammer/Pine Bunting (only ten such records in eBird, so its potential potential is not fully realised), as it's not a pure Yellowhammer but, depending on location may be considered a Pine Bunting hybrid, so (somewhat) in a grey area--much like (I believe)
Larus sp. would be used for a bird that lies between, e.g., Herring Gull and Caspian Gull, but not all features could be examined confidently enough to confirm a hybrid, or the way one would use redpoll sp. for unassignable Redpolls.
ebird.org
I like the way Zulfu Farajli uses the label in an area in which both are reasonably possible.
It would have to be reserved for special cases in areas in which Pine Bunting ancestry is reasonably rare, as it is a slightly different case than the above): that is, when the bird's appearance suggests that there is a chance that it comes from the hybridisation zone (which may be slim, but the possibility can't be casually proved or disproved).
Still, eBird assigns slashes to the most probable species in a given area (from what I remember reading): Yellowhammer, in this case, so there should be no harm to the data model.