It's really amazing work, and the knowledge and skills NASA and collaborators learned from doing that are invaluable.
I honestly think for the next decade or two that NASA should double down on propulsion science, energy science, and robotics. Instead of trying to send a few men and biospheres to Mars, lets send a greater quantity of increasingly advanced robots/AI on increasingly challenging missions to study Mars and the Asteroid Belt.
Use NASA's limited budget to advance theoretical and applied AI, robotics, and new propulsion/energy generation technologies for industrial use in space, with the ultimate goal of creating a fully robotic space mining and refining industry and supply chain back to Earth.
Keep the ISS of course, as a relatively less expensive way of studying the effects of low-G on the human body, space agriculture, and human sustainability in space. But for anything beyond Earth orbit, let robots prepare the way for humanity.
If NASA had an unlimited budget, then by all means do it all. But unfortunately that's not the reality right now. The robotic missions of the last several decades, from Mars Rovers, to Hubble, to probes of asteroids and the outer planets, have had such an amazing track record and ROI that they're worth doubling down on for the foreseeable future. The rapid advance of both AI and robotics is also expanding the mission capabilities of this technology, and its worth pushing that envelope hard for another decade or two.
Finally, given limited budgets, we can send a greater quantity of robotic missions than we can manned missions. Quantity is an important factor in quality, as I think the tech world has learned in the last decade or two.