To your point: "why would funders - such as foundations - who are ostensibly primarily concerned with the social return on investment balk at the diffusion model if it is better suited to the innovation which they hope to spread?"

As you state, I don't know that foundations would necessarily balk at funding innovations built to diffuse per se, other than the fact that foundations are probably more comfortable funding "organizations" than funding "innovations" in the first place.  There are at least two reasons why this is the case.

First, foundations (and private capital) are more likely to fund proven successes - either (A) people and teams with a track record of success (B) organizations that have already achieved some level of success with their current endeavor and need funding in order to grow (i.e. scale).  If you spend any time talking to traditional VCs, they will be quick to tell you that they fund teams rather than ideas.  Although foundations may be less explicit, I imagine they are often looking for the same thing (a quick look at the criteria Ashoka Fellows of Skoll Foundations Awards confirms this).

Second, and I think this is the really the key issue, is that it is hard to measure the impact of dollars spent on enterprises whose impact comes primarily through diffusion rather than proprietary growth.  Foundations, as much as entrepreneurs, want to be able to "take credit" for their success.  Ego aside, there are many altruistic reasons for taking credit for success (especially given the point just made about funders being more likely to support entrepreneurs/organizations who have already proven themselves to be successful).  What I am getting at here is that funders are interested in backing people/teams/organizations that are likely to succeed, rather than support ideas (even if widely replicable outside the organization) at organizations that are likely to fail because it is much easier to demonstrate their impact at the former.

From the perspective of pitching foundations (as opposed to private capital) for funding for diffusive organizations, what I think is needed are ways to clearly articulate and measure impact in ways that are proprietary, even if they do not involve than proprietary "growth" per se.  In other words, the organization needs to be able to show how, and to what extent, it is directly responsible for diffusing the innovation.  Sean's Steal this Idea! blogpost compares Nurse Family Partnership (NFP) and Homeboy Industries.  Sean describes how NFP replicated its program all over the country and "now offers services in 28 states and have over 16,000 families enrolled in their program."  In comparison, Homeboy Industries has refused requests to replicate its gang intervention program from Los Angeles to other cities, but, according to Sean: "they do act as a model for other programs and help other programs get started."

Just looking at Sean's own descriptions of the two programs betrays the problem:  NFP gets credit for over 16,000 enrolled families while Homeboy Industries' claims only the rather vague impact of "helping other programs get started."  With foundations and NPOs focusing more and more on impact and accountability I think the reality is that most funders would prefer to contribute their scarce resources to the NFPs of the world rather than Homeboy Industries for just this reason.

Instead, the onus will be on the entrepreneurs to think about how to approach diffusion in a way that is directed and measurable rather than the sort of vague, serendipitous diffusion that Sean describes for Homeboy Industries.  For instance, Homeboy could create a standard curriculum or standardize other components of its program in a way that could be copied and reused by other upstart programs in other cities.  With just a little work, Homeboy could track the number of organizations and students using their curriculum and could claim, for example, the "16,000 former gang members who have been trained for new jobs using their curriculum in 28 cities."  This certainly helps a potential funder more fully appreciate how the organization can achieve a large "footprint" even without achieving proprietary scale.  I think it will remain up to the entrepreneur to think proactively about how to design organizations for such "purposeful" diffusion and how to effectively measure the organization's diffusive impact.