But this analysis makes some assumptions, which are mostly forgotten in this sound byte world we've built. The assumptions are that the goal is scientific results, and that only monetary costs are important, and probably others. The Mars rovers have been doing a great job at geology for the past five years. Totally off the wall fabulous. But in a typical day, they move several feet, take a few pictures, and maybe move the arm to some object and take a few measurements. Great science, but a human could do that sort of work every few minutes for hours every day she's there. And at the end of her many months mission, she'd have advanced the geology of Mars science by a century of what the rover could do. Sure it would cost more. Sure, the same money spent on robotic missions would get more science. But the robots would take a century, even assuming lots of simultaneous missions.
Jack Schmitt (Apollo 17 geologist) was off the wall amazing. It's really too bad that the next couple flights were canceled, and they waited so long to put a geologist on the Moon. There's only so much geology training you can do with astronauts. And that's why NASA is training geologists, biologists, etc., as astronauts. They've already got their decade+ of PhD grade research in some specialty under their belt. This is depth you don't get in sound bytes.
The assumption that the only reason for going into space is science is also suspect. There's also the adventure and exploration. You can't think clearly about a Mars colony if you haven't put people there. The Earth is the cradle of civilization. But you can't stay in the cradle forever.