to some degree, yes. HOW to scale up is a big problem. you are correct - there is no guide or set of standards. so, that being the situation, the question is how to deal with it.

when faced with high uncertainty of future results, managers are forced to make assumptions. but, are their assumptions correct or incorrect? how can we tell?

the answer is that we can't tell, before hand, which assumptions are correct or incorrect. if we could, we wouldn't have failures - either with BOP ventures or with official development efforts. so the question then becomes, 'how can we determine if these assumptions are correct or incorrect (and avoid failure/ create success)?' what do we do?

to handle this situation, some in the business + academic world suggest using "discovery driven planning", "assumption based planning" or "critical assumption planning". these techniques all focus on the same things:
1) identify your assumptions
2) design inexpensive tests to validate or invalidate your assumptions
3) change your assumptions based upon your tests
4) return to #1 as necessary
5) go for it or shut down based upon your new knowledge

failures are frequently the result of trusting assumptions that were actually incorrect. when Easterley writes about 'Planners', this is a big part of what he's talking about. ...to the degree that we can identify the assumptions made and how they actually panned out, studying failures can work well. studying successes can work well, too, and for the same reason.

but, you posed the question, "HOW do we scale up small, localized experiments?". the answer may sound strange, but i believe that the way we discover HOW to scale-up small, local BOP experiments is to design small, inexpensive assumption-validating experiments aimed at that very issue. ...think of it as 'organized searching'.