A laboratory on Mars beats MSR. Surely?
The prize is not a box of rocks in Houston. The prize is answers. Age, chemistry, organics, context, and whether Mars ever held anything alive.
Classic sample return has a long relay. Collect tubes. Land another ship. Fetch the cache. Fire a tiny rocket. Rendezvous. Fly home. Build a special Earth lab just to open the box. NASA's own timelines slipped into the 2030s. Costs blew out toward $8-11 billion. The existing NASA-ESA architecture then lost support in early 2026.
Even the cleaner commercial ideas had the same clock problem. A 2031 return is still late if Starship is already on the surface with Optimus androids, and people not far behind.
So put the lab on the ship.
Staff it with Optimus androids first, people later. Analyse the Perseverance cache on site. Then keep sampling.
That wins on time:
You do not wait years for ascent, Earth entry, and quarantine. You start when the ship is down.
It wins on volume:
MSR is a handful of cores. A Mars lab can go back to the depot, drill the next target, and run the follow-up the same week the first result looks interesting.
It wins on context:
A rock in a glovebox has lost the ground it came from. On Mars you still have the outcrop, the dust, the ice, and the next sample 20 metres away.
It wins on failure modes:
Classic MSR stacks landers, helicopters, a Mars Ascent Vehicle, an orbiter, and a receiving facility. One failed hop and the whole set is gone. A lab on a ship you were sending anyway can run the test again tomorrow.
Yes, Earth has huge instruments. That used to settle it. It does not anymore. With the right equipment on board, what can we still only do here that cannot be done in situ there?
We get attached to holding the rock ourselves. That is a feeling, not a science requirement.
Landing site is the real catch. Starship is not currently aimed at the Perseverance depot. Fine. Land closer, or send androids with a vehicle to collect the tubes. Crude. Still cheaper than standing up another national relay. Or send a Starship with a lab to several places - there will be enough.
If you still want rocks on Earth, send some home later on a returning Starship. Analyse first. Ship later if you must.
MSR moves the pile to the microscope. A Mars lab moves the microscope to the mountain.
Samples will come home one day. By then the interesting work should already be done on Mars.
A lab wins, surely?