> Does anyone have ideas on how that may be achieved - what a correct incentive structure for research might look like?

Perhaps start with removing tax payer money from the system.

Stop throwing good money after bad.