> Some science really can't be done as well or cheaply by robots alone.

I'm struggling to imagine how a "lab full of equipment, life support, beds, bathrooms, comms, and return rocketry" doesn't ($-wise) translate much better to "many, many more teleoperated robots with miniaturized labs operated from orbit or Earth with frequent sample return round-trips".

Boots on the ground is a moral victory, but I don't see the science victory until you get enough equipment / personnel there to get a thriving ecosystem of thought in-situ. Not to mention: Astronauts are rarely the scientists that study the place they are going.

Now an ISS around Mars? That makes a lot of sense to me.

Humans can get a lot more work done, just by their dexterity, flexibility and ability to learn and adapt. For robots, even simple tasks have to be pre-planned, in this case sometimes years in advance.

It is conceivable that you could build a robot/machine that can collect one type of sample and do one type of analysis cheaper and quicker than a Human. However, human astronauts can do multiple types of these, maybe even adapt them on the fly.

for this, teleoperation from orbit seems reasonable. It's not been tried however.