I've written here before with some back of the envelope numbers for a hubble 2.0 + starship as the LV. It gets more interesting if you set up an assembly line f

I've written here before with some back of the envelope numbers for a hubble 2.0 + starship as the LV. It gets more interesting if you set up an assembly line for cameras in a tube like the hubble to spread the costs out, and use newer guts for the sensor payloads and on board compute. Make some assumptions about the telescopes, such as being okay to fail since we have so many up tWebb. You could even do experiments in computational photography like the Keck or the VLA to overcome the light collection limits of the mirror size. I forget the numbers I came up with but it's a lot of orbital sensors for the cost of one hubble program and like an entire space station with sensors for the James webb.