"elicit subjective evaluations from a large set of experienced entrepreneurs and executives"

Sounds like they're including evaluation by VC-types, not just industrial or scientific expertise. Point being that R&D intensive work is more predictable in general, and VC predictive ability may be overstated.