You'd have to overhaul the funding system drastically.

Measuring scientific output by publications and citations is useless at best, but it's easy so that's how you're measured.

Writing grant proposals is 95% useless bullshit, there's no useful content in the proposals, but it gives a false sense of objectivity and competitiveness, so that's how you're funded.

Thing is, most of the world operates like that. Corporations measure useless KPIs and demand empty reports. There's an entire caste of administrators whose entire existence is founded on this overhead to exist. I don't see a way to change that without a very very serious disruption (that is, a major war, not a startup).