Next, task NASA with a “Inhabit Mars First” priority mission. Minimal on-orbit facilities, focus on using what we already know, full stack testing (cheaper, faster, and more effective when using simple components). No moon work. Why? Because, despite common belief, it is more expensive to to Mars via the Moon (or, frankly, via LEO) and most of the truly important hardware is not reusable for Mars — though the opposite is somewhat true, Mars hardware can be modified for Luna use. For projects and programs not directly supportive of the Mars program, the total expenditure must come in at no more than ten percent of the total NASA expenditures/budget.
It’s cheaper than that.
Five years ago a Dutch group asked an international group of aficionados who would volunteer for a Mars mission under the assumption that it would be a one-way trip. IOW, no way home. You live there, you die there.
According to their math, 2/3 of the cost of any Mars mission (via Luna or not) is getting people home. If you assume there’s no return trip, the costs drop like a rock.
They had far more volunteers they could ever have dreamed of. All the necessary skills were present in ample numbers.