I absolutely like your logical train of thought, but it isn't likely given what we've seen historically. We on ENWorld and gaming internet in general live in a real bubble where we see very clear repeated patterns of how folks buy things, such as on D&D Beyond. But for D&D, the audience is vastly bigger than our bubble. There are folks constantly walking past store aisles, coming in to gaming stores, shopping online. D&D is a game with global shipping and distribution schedules. The scale of it is enormously beyond what we see. Most sales, historically, come from outside the patterns of buying we see.
For sure, digital sales are a new thing. But in the vast D&D market, the number of folks who have even been to the D&D Beyond website is likely very small. When I would visit gaming stores around the country in the first few years of 5E, maybe one in six D&D players had been to the D&D web site. Fewer than that were engaged in anything forums/social online.
While everything you mentioned is true (digital sales, FLGS sales, bundles) and a factor in sales, it is inconceivable for those factors to so profoundly impact the BookScan side of things (we haven't seen such a change for previous 5E books, even during the D&D Beyond digital era). That's why folks who understand BookScan tend to think that the numbers being reported are inaccurate for some reason.