This means you can spend more fuel/be less efficient overall for a much more comfortable acceleration profile.
In terms of safety, starship will fly hundreds, perhaps thousands of times before people fly on them. They'll get there.
In reality a Starship flight would probably sell few very expensive seats instead; just like Boom's aircraft only has 50 seats. Which would make it worse than subsonic flights in terms of fuel used, but still competitive with supersonic.
For a flight time between 20 (minimum) and 40 minutes (maximum) to anywhere on earth, this volume could comfortably fit 1000 passengers per flight.
At 2 or 3 times the price of a regular ticket, their offer would be a direct competitor against business class tickets, if they can reduce enough the risk, and handle logistics of rapid reuse.
Also, starship has more pressurized volume than a 747. You can fit a lot of people.
Realistically, P2P using giant rockets is a nothing but a pipe dream. Just straight forward things like weather proof this. The latest F9 launch had to be scrubbed two times due to unfavourable weather at the launch site, for example.
Now multiply that just that weather-risk by two, since the landing site also requires good conditions at the same time.
Then there's the "little things" like airspace closures, launch permissions, boarding procedures, the requirement for personal pressure suits (simple masks won't do in case of cabin pressure loss), etc.
The idea sounds so easy on paper, but there's a lot of good reasons that in over 60 years of crewed space flight, the idea hasn't even been demonstrated.
Now for the launch platform, it will have to be built about 30km from shore, and could be accessed by high speed train, the travel time would be 15-20 minutes from the shore
Fuck dang, time to cycle residential proxies.
That said, see sibling comments about fineness. In addition to being fine in the shiny variety, starship is a portly fellow. It will be able to land in anything less than a hurricane.