Thanks much for this post; it's an important contribution. On the 'necessary set,' you have researcher and implementing agency inputs. But what other stakeholders might fall in the ideal set-up? For one, the implementing agency may not be the decision-makers about scaling. For two, the example you give is about other organizations and banks. How might you have better understand demand and timelines for evidence from other such stakeholders during the formative and design phase?
Also, I think your post raises the importance of having a good monitoring plan and system in place.
Suvojit and I have a post over here (http://blogs.worldbank.org/publicsphere/relevant-reasons-decision-making) that might provide food for thought about the decision-points you discuss for scale-up. Always happy to hear your thoughts on it!
Thanks again for a good post.