This is not really true. There are plenty of ideas that don't involve a significant R&D phase (i.e. consist of more or less the implementation of the idea rather than the research-based refinement of an idea) which can immensely valuable.

Plenty of examples exist such as Uber, Basecamp, etc. Although the success of these ventures is dependent on fundraising and marketing, these are certainly not the only factors and these factors apply to research-heavy startups as well.

"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.