This was kind of a fun, interesting idea I've had for a while, and one that definitely warrants further exploration. Using a capo on my Kingston Super 5 and making just minor action adjustments, I played the same riff to compare effective scale lengths of roughly 35", 33", about 31", and about 29.5".
I made the conscious choice to stick to fingerstyle playing and using longer notes (along with a faster passage) across all 5 strings to try to highlight the variations in timbre as much as possible.
Surprisingly, to my ears - and headphones/speakers are definitely necessary to really pick out the differences - there isn't as much tonal change as I expected; there was considerable difference in feel though, that's for sure. And comparing the two extremes of lengths certainly sounds different. I did slap right at the end at 29.5", and that sounded weird, haha! String bends at that length can exceed like a perfect 4th though, so that's fun. Getting close to baritone guitar territory there.
One thing I noticed that now seems kinda obvious: if the pickups aren't moved backwards with shorter scale lengths, they're effectively moved forwards. That might make the most noticeable timbral change. Maybe? You tell me.