I find the issue of different apodizing implementations and "hardware capability" rather "interesting" (which might be a polite way of putting things).
I'm sure I'm not the one one wondering whether the information we have on "different apodizing implementation" was outright false, miscommunicated or misinterpreted. For instance, was a comment about the overall product's "hardware limitations" being applied to, or interpreted as being about, its apodizing in particular? Meridian does now seem to be a company that delivers new product rather than updating old ones, so I wouldn't be surprised that there are marketing forces at work to convince customers to upgrade.
We know that the 808.3 sounds better than the G08.2 because there is so much more care and engineering involved in the design and circuitry; and the "implementation" of the up-sampler (chip used, position on circuit board, colour of the case

) might be different, but do these products really have different upsampling FIR coefficients? I think that it's likely that once Meridian settled on its preferred apodizing algorithm after (presumably) a *lot* of listening tests, it implemented the same FIR filter in all products. Evdberg's results might suggest the same.