>>13676072A lot of delays with JWST is simply a result of technology that did not exist and needed to be invented in a research lab, and then scaled out into manufacturing, etc. For example, the camera sensor in JWST was basically impossible until fabrication process got down to around 32-22nm. That took nearly a decade. Then there's the gold coated beryllium mirrors:
https://www.jwst.nasa.gov/content/observatory/ote/mirrors/index.html | JWST has 18 of them, those also took a decade.
Its true that there's a lot of eagleworks/skunkworks with JWST, but all of that had to happen, because of the kind of work it was intending to do. Take LUVOIR for example. That's Hubble's true successor, and even with all the experience from JWST factoring it, building that observatory will likely take another decade or more--as the sensor suite necessary for its science missions more than likely do not exist and will need to be invented for it.
Even with Starship's upmass capability, there's a large amount of creation involved that will cause delays. Hell, LUVOIR is expected to have 35 mirror segments. They'd basically use the same system of Beryllium mirrors because they're strong, light, and flexible while being able to be polished to micrometer precision. On top of that, those segments are ~40% bigger than JWST, so the actuators, the frames, everything will need larger variants, etc. There'll likely also be ESA/JAXA partnerships + potential Artemis Accords partnerships for science instrumentation on that. That too will drive up costs and delays.
If JWST cost $10-12Bn for launch, I'd expect LUVOIR to cost ~$23-25Bn to launch and get put into orbit ~2035-2040. The only way LUVOIR gets completely bypassed is if by 2040, there's a sufficient lunar and martian presence with budding mining industry on both worlds as well as the belt with common transport networks between all three locations and Earth that allows for the creation of LUVOIR class or larger observatories in space.
