Take photos and retain specimens - best of both worlds?

I have previously written about the way John Bridges has greatly increased his site lists by retaining specimens that he passes on to me for ID. There are a few others who do the same. Yesterday I worked through a batch of specimens collected by Martyn Hnatiuk who is based in South Wales. That was highly instructive and useful because Martyn managed to collect several species that are very rarely recorded and would not have been identified from photographs. Three stand out above the rest:

Cheilosia carbonaria - a species that is regarded as 'Nationally Scarce' and which I have not seen in ten years! This specimen confused me a great deal because the antennae were quite strongly orange but the wings were smoky. At first I wondered whether it was a new species but upon checking Van Veen it ran very easily to C. carbonaria. This is an object lesson in the need for care when using what seem to be reliable characters. In Stubbs & Falk, C. carbonaria sits in a group of species with dark antennae and this specimen refused to go that way. So, we would never have achieved a reliable result from a photo and using Stubbs & Falk. It shows that there are always the exceptions that prove the rule!

Eupeodes nitens -  I have only seen this one a couple of times, several decades ago, and there are very few reliable recent records! Again, it did not wholly follow the key in Stubbs & Falk, but readily dropped out using Van Veen. This one would never have been done from a photo because the critical diagnostic characters are on the sternites of the abdomen.

Trichopsomyia flavitarsis - a species that is closely associated with very acid sites and which we very rarely get records of. At one time I used to see it pretty well whenever I went to the north-west (or at least that is my memory) but I've not encountered it for a very long while so it caused me quite a lot of confusion at first.

So, I wonder what else is out there that more recording might turn up? Martyn got a fair selection of  species that we would not have managed to ID from photographs, thus greatly improving the resolution of his data. John Bridges has done much the same. Perhaps others would like to do something similar?