I got interested in Starship from a project management point of view, as an example of applying an iterative style adapted from software development to a large-scale hardware project. On the plus side, it makes possible, for example, going from transpiration cooling for reentry to a tile heat shield, a change that ArianeSpace or Blue Origin might have struggled to do without a complete redesign. On the other hand, it makes it possible to ignore technical debt from incremental changes that accumulate until you find the payload is half what was spec'ed.
The biggest challenge SpaceX faces are that key milestones are inter-linked: It has to be cheap; it needs to have a big payload, over 100 tons; it needs to be really rapidly reusable, like restack, refuel, and relaunch. If any one is incomplete the others are out of reach, and it's only good for big payloads to LEO.