The idea of making fuel on Mars was popular enough to have the calculations be part of undergraduate thesis' over 20 years ago, back when I was in college. I bet that people far more serious than Aerospace Engineering undergrads have been coming up with far more detailed versions decades before that. But those plans are no panacea: The fuel creation takes time, so you are either sending a mission first, just to make fuel, hoping it all goes well, and then send a second mission later, or you might end up having to keep your rocket in the surface of mars in good shape, still able to come back, for quite a bit. You also have to consider general timing: It's not as if efficient trips in either direction can be done any random day, of any random year, just due to orbits.

So yes, the mars fueling idea is compelling, but let's not assume that the people doing this for a living, and proposing this plan, are just worse at this than random undergrads.

Can we "just" build a Mars space elevator? It should be free of a lot of the concerns an earth space elevator would have - no terrorists, no populated areas, rather less gravity.

I've heard that this could be done with current materials science. Same for the moon. The problem is that it's a solution without need for now. It's somewhat like building a cross continental railway before San Francisco was founded. One day hopefully we'll need something like the elevator, because that implies lots of other interesting things would be going on.

The problem is Phobos, which orbits closer than areostationary orbit.

Could we just tether Phobos with an elevator?

Since Phobos isnt areostationary, the base of your tether on mars would have to move around for this to work. Additionally, you would have to constantly adjust its length. Therefore, boarding the elevator would be more difficult.

Or set the height above the highest obstacle on the orbit and have some variant of planes or wilder mechanisms get the cargo up to that dangling tip.

The main difference is that SpaceX operates very efficiently, while most public institutions (like JPL, which currently develops the lander, apparently while being ill-prepared for the task) are highly inefficient.

A striking example is the large amounts of money NASA spends on SLS, which is in most places already an extremely conservative vehicle which reuses as much as possible and keeps any innovations at a minimum. With the JPL lander the situation could be even worse.

As we just saw in the news this past week, the jury is still out about the wisdom of building and operating complex, dangerous missions "very efficiently" and with "innovations".

The Falcon 9 is very efficient and innovative, but also one of the most reliable rockets on the market.

Yeah. By itself Starship is indeed very dangerous for human spaceflight (no launch abort system, very risky landing procedure) compared to other designs except the Space Shuttle. But the reliability can be optimized while doing a lot of unmanned satellite missions.

We are hundreds of successful starship flights away from it being human rated IMO.

Agree 100 percent. But also why return the samples at all? Just send automated lab equipment to analyze it.

The article says the utility from sending the samples back is indeed limited, at least compared to the cost, which would likely exceed the (already astronomical) JWST cost substantially.

No amount of automated lab equipment is as flexible to work with as samples on Earth, for obvious reasons.

There's no such thing as automated lab equipment that can analyze it. Even the space station has to send air and water samples down to Earth for anything beyond very rough analysis. You need technicians and a cleanroom environment to do this work.

How about 10 billion to NASA instead of privatizing research with public money.

Because NASA is absurdity inefficient when developing technology. That's why SLS is such a problem child, and why they abandoned their new space suits, which were already years in development and consumed billions of dollars. They were recently outsourced to a private company which is expected to do it in a fraction of the time at a fraction of the price.

Because NASA hasn't done anything useful at a reasonable cost in decades.

I'd rather a somewhat inflated cost to make the research public. Also private equity doesn't mean it's not wasted. 37 billion dollars for a forum.

Musk really living rent free eh?

How many rockets do you think NASA themselves have R&D'd?

A lot of those 10 billion would go to aerospace contractors either way. It's not like NASA would spend all that money in-house.