For sure. It's hard to keep in mind the many many considerations this topic carries. The sales of VR are awesome compared to most historical sourcebook/setting sales, especially for that type of product.
I wish Ray could share what he knows about the holes in the data. We all know there are holes... but can those holes be characterized?
Some possibilities (assuming Ray is really saying that the BookScan data isn't indicative or relative Van Richten's sales, which may or may not be what he is saying):
1) D&D Beyond is huge. Maybe DDB has huge sales for something like VR, dwarfing paper sales... the latest Hasbro Q3 earnings has enough information to strongly suggest DDB sales are not showing huge growth (As we discuss in this Wednesday's episode of Mastering Dungeons) and maybe maybe that they also aren't enormous compared to other sales types. But it is really hard to tell.
2) Bookscan is Weird. Maybe Bookscan wildly diverges from other sales sources for some products. Like, maybe VR sold tremendously well at FLGSs. Maybe there isn't a good correlation between BookScan and other sources. But it would be pretty surprising for that to be the case.
3) Maybe the BookScan % is way off. Maybe BookScan is a far smaller piece of the pie, and the other pieces are big enough that what sells well on BookScan just isn't that huge a factor. VR might have done poorly via those distributions but that is dwarfed by how it did at stores and other places. This is perhaps the one I see most possible, but the little data we do have shows BookScan being fairly significant. If we could get more data, such as D&D announcing the number of a product sold in a particular time period from all sources, then we could get someone with BookScan to check and find that percentage.
There may be other possibilities, but I am not sure what they would be.