I don't think Linux has an equivalent of Apple's vision API, and if it does I guarantee it's not as robust and isn't baked-in to every Linux distro (the way Vision is baked-in to every Mac released after X date).

That alone will likely prevent this from just being a "convert to Linux" vibe session ... which is unfortunate, as I would LOVE to have this on Linux.

Yep, this is the bottleneck. Otherwise I would have done the prompt myself.