Certainly nothing is permanent, and specimens have sustained their proportion of losses over the centuries. But at least we can talk about their track record in terms of centuries. What is the track record of Internet web sites over the last 3 decades?
Not wishing ill will to iNaturalist or any other such endeavor, just that with existing track records, the jury is still out on our digital infrastructure. I would be hugely and joyfully surprised if I could come back 300 years from now and find all of your and my iNaturalist observations and IDs still serving the cause of biodiversity in that world. I wouldn’t be here if some part of me didn’t think that possibility existed. But would I bet life or treasure on it right now? Hmmm… no.
I am certainly a huge proponent of the “photo-vouchering” work that I and all of us are able to do here on iNaturalist and in other venues. Its value cannot be overestimated, especially in situations like you mention where specimen vouchering is impractical, or would be “overkill” (pun intended). My point is that photo-vouchers versus specimen vouchers should not be a zero-sum proposition. I will push back against any notion that the value and need for specimen vouchers decreases in proportion to the creation of diagnostic photographs and DNA barcodes. (Maybe that is not what you are suggesting, but it can be heard that way…) In fact, I would say it is exactly the opposite. Each contributes value to the other.
Trying to bend back to the original topic of DNA barcoding…
Taxonomists continue to learn new things all the time about what pieces of evidence are critical to understanding species boundaries in nature. If one doesn’t happen to capture the right evidence in a photograph, or in a DNA bar code, those pieces of evidence can’t be reconciled with new knowledge. But with a well-prepared specimen, you have the whole organism, including the DNA, to go back to and re-sample.
One recent example in plants – over the past decade, there has been a huge leap forward in understanding the difficult genus Boechera (formerly in Arabis) of western North America. It was discovered through chromosome counts and careful study that one could tell the sexual diploid versus apomictic triploid status of a specimen using pollen size and shape. Once species boundaries were better understood on that basis, it further became apparent that basal lobing of the mid-stem leaves is a very good corollary distinguishing feature among species. Until these discoveries, I never thought to photograph the mid-stem leaves of Boechera plants (much less the pollen, which I still don’t!). Without all of the existing specimens (and especially, of course, the type specimens, some almost 200 years old), it would never have been possible to reconcile this new knowledge with the previous taxonomy for these plants.
I think we are both valuing the same two things here. I just don’t see the value of one as being in the position of diminishing the value of the other. I see them as enhancing each other. Specimens will remain just as important as ever, and the value of photographs and DNA barcodes will continue to increase, and to augment the value of the specimens on which their very identity ultimately rests.