Iterating on Hubble would certainly be a bit more realistic, the Nancy Grace Roman Space Telescope - due to go up later this decade - can be considered somewhat

Iterating on Hubble would certainly be a bit more realistic, the Nancy Grace Roman Space Telescope - due to go up later this decade - can be considered somewhat of an iteration on Hubble in that it is also based on the same chassis with modernized optics, although IIRC it focuses on a different part of the spectrum. Similarly, many proposed successors to JWST also iterate on the basic design and technology for JWST to include more mirror segments. However, I suppose you're referring more to a silicon valley style iteration, which I think isn't worthwhile yet since ground based telescopes are already close to parity with Hubble and are cheaper to iterate on.