If I'm understanding your question right, that's exactly what we're seeing at the end of the video (normal upresed SD compared to AI enhanced).
it's likely i'm not making myself clear enough, so i apologize in advance. let me try one more time.
the normal uncompressed SD clips in his video, before they are put through Topaz, are 720x576. after they are put through Topaz, they are "enhanced" as well as now upscaled to HD as 1920x1080.
what i typically do with VX clips that go to youtube, is use FCPX to import that raw SD clip into an HD 1920x1080 timeline. this obviously upscales that SD clip into HD, which is then exported.
my question is, if we start off with a single raw SD clip and run that through the Topaz, and take that same raw SD clip and import into FCPX HD timeline as i described above, would we be able to see a difference when comparing the Topaz vs FCPX clip?
at the end of his video, he compares the raw SD vs the result from Topaz, which is HD. obviously, we are going to see an increase in quality. i'd be interested to see if there was a difference between the Topaz clip, and an upscaled HD clip from FCPX.