This may well be varroa. Last summer, we pushed the varroa edge in our app testing and we subsequently took our 'before' end of Dec losses. We knew the risk, but wanted some answers.
As per the tens of thousands of dying colonies (presumably they lost the bees not the boxes) I'd rather see the results of sampling and analysis of dying/dead bees, if any.
Varroa should be in the samples in high numbers. What about Nosema? First year without fumagillin? Send samples to USDA or Dave Tarpy for PCR for viruses,Dave Wick for general virus scans. Talk to Dave Wick about the bacteria he's tracking and can now look for with new instrumentation. Do some chemical residue testing - it's a possible, but improbable factor on this scale.
One hopes all of this is being done by the affected beekeepers.
If I took a 50% loss, I'd want to be sure about the pest, disease, even pesticide status of the best colonies compared to the struggling and to the dead colonies. The costs should prove to be far cheaper than making the wrong call, basing corrective action on a guess. Varroa is possible, probably probable. But what if it's something else? For tens of thousands of lost colonies, set up an organized, properly stratified sampling and testing plan and look at all of the factors - biological and chemical. Let's stop the guess to assign cause. Let's also not wait for someone to come out - days or weeks later. Whatever factors may be involved may have disappeared, perished, or degraded too far to detect or to identify.