The "quite soon" is no earlier than late November due to licensing issues (such as that the booster return doubles the sonic boom zone).
Some points to consider:
I am not expecting a successful StarShip landing on Mars by a launch in 2026! Maybe he could get a StarShip there in that time frame, but how does he intend to slow it down and land it? Mars' atmosphere is not nearly as dense as Earths, so braking by atmospheric drag is not going to be anything like an Earth reentry for StarShip.
The reentry profile is almost exactly like an Apollo reentry was, due to the similar Mars atmosphere vs Earth stratosphere pressures.
And, if the rich are not signing up in large numbers, how is a Mars colony going to be profitable in anything like the medium term, before it can become self-sustaining and produce profitable exports?
The estimates of willing colonizers that can afford the transfer, at the cost of a typical US home, should suffice. The "billionaire" stuff is a dud meme.
The economy of an exporting Mars society after the settling economy is dubious. It is not analogous to Earth colonies that were established because they knew they could export raw materials.
Perhaps exports such as software production and technology licensing will contribute to expand the colony economy, who knows!? But the potential colonizers don't seem to be bothered, and notably Earth does well without exporting to other planets.
He has one very hard problem with launching people to Mars. There is no way for them to return to Earth. After Starships lands on Mars, its fuel tanks will be empty. He will need (1) a fuel factory and storage depot on Mars so as to be able to refuel Starship and (2) a launch pad.
The current loose plans show robotic assembly of refueling plant and return pads. The return option will be there before humans are sent. But they seem to expect that few of the later colonizers will want to afford it.
I agree. Whilst I think Starship would be useful for getting cargo to Mars when it comes to getting people there you want the journey time to be as short as possible, due to the health risks of a) prolonged weighlessness and b) cosmic and solar radiation during the voyage (once on Mars the radiation dose can be reduced by building shelters covered in Martian regolith). Which points to the advantages of using nuclear thermal propulsion for the human transportation part of the project.
Like billionaires, the infinite improbability nuclear drive is a dud meme. The transport system is chemically driven.
The weightlessness of a 5-6 months Hohmann transfer is far shorter than the longest ISS stays, and after the 2-3 g descent the stay will be 1/3 Earth g gravity acceleration.
The cosmic ray exposure is not much different from ISS trips. It is mostly Earth atmosphere that shields. Away from the planet shield the douse will be roughly doubled, so they will incur some cancer risk in later life, like people who likes red meat or alcohol (or both).
For solar CMEs and wind - which is the harmful part that the crew can be exposed to - they will put the water tanks between Sun and crew.
They likely want to have ship recycling and prepared martian greenhouses before sending people as well, since they else will have a mass problem.
#1. There is no way to abort a launch, If anything goes wrong everyone dies. It will be really hard to get approval to launch people to space in Starship
#2 Starship can not leave low Earth orbnit until it is refueled in space and this requires 10 or 12 tanker flights.
There are not problems, these are features.
Consumer vehicles don't have clumsy and energy inefficient "abort" systems, specifically airplanes have never had useful such. (Parachuting from a crashing plane is not an option.) And the transport system is set up to approach airplane system safety. (If they get there is another question.) The most unsafe transportation where I live if you count time spent is city walks, cars hitting walkers means many orders of magnitude higher risk than airplane transport.
For most of the launch the Starship upper stage in effect serves as an accidental abort system, as they have already demonstrated (hot staging separation). Not a good one since it may fail an uncontrollable booster if all engines are on. But there is no way "there is no way to abort", since there is. (If it is approved is another question.)