With a telescope integrated into Starship, you could potentially land the whole telescope from orbit and do repairs and upgrades on the ground. Whether the main mirror can withstand launch and reentry is a question though
That also depends on how Starship can actually open which they've been pretty vague on for larger payloads. It seems difficult for it to open in a way that gets the bottom half of the front of the ship out of the way (because of the heat shield). If you can't fully clamshell that open you might still need expensive and delicate origami to unfold the main mirror and any sun shields to get them a clear view of the sky.
That's setting aside the costs of actually running it too as brought up in the article. Maybe that can be brought down but there's necessarily going to be more complexity and more specialized workers required to run a space telescope vs a ground telescope. It'd be interesting to see where the extra costs come from; the specialized people monitoring the satellite, downlink time, etc. Some could get brought down but seems difficult to make it cost competitive with having a similar telescope on the ground.
EDIT: found a counterargument: https://www.scientificamerican.com/article/elon-musks-starsh...