Specifically a manned flight to mars, which before depended on the cyclic political will and state of the cold war to be funded. There are also other uses for starship like 30 minute suborbital flights between new york and london. Maybe it's easier to see if you look at the limits: given an infinite amount of money and labour, we wouldve seen a manned mars landing in the 1980's perhaps. If on the other hand funding was finite and reaching mars cost 100% of world gdp, it would never happen.

SpaceX launch cost is roughtly half of competitions for same payload: 40mil USD vs 90mil (Ariane Space).

40mil is far more affordable than 90mil, and they are promising lower prices in future.

Starlink becomes economically viable.

Since the satellites are mass produced each one is comparatively cheap. That means losing them is an option and SpaceX can launch them using parts that no other customer would want (first stage boosters that have been flown 4 or more times already). Their main costs are the second stage ($20 million), payload adapter and fairings (which are simpler than the normal $10 million version), fuel ($300k) and ground support etc ($1M?). That puts you in the range of $30 million. A regular launch would cost twice that, and any competitor easily twice of that again. So it's a 4x launch cost reduction, and a lot of the remaining cost can be written off as R&D as it's exploring the failure points of reused boosters, eventually allowing them to put regular customers on more heavily used boosters with confidence.