12 Port-
As long as you don't develop a hexa-polar disorder, that is a very good point about minimalizing costs, and where those development costs are usually incurred. Production tasks cost money- plain and simple.
Once you get the pencil on paper and rough-order magnitude of the project, there should be a decent amount of support from here and elsewhere.
In reviewing several material change projects over the past three years, there is one huge factor that is probably on your side that you didn't mention- time. In every one of those projects, more often than not the desire (or requirement) to have a component ready for 'plug & play' application on a set date accelerated each of the duties of the team members. This can create more than one "snowball" to keep in check.
I already know my answer to this question, but it is also thought provoking: At what point would a cylinder head 'redesign' or 'modification' totally reconfigure the engine's specifications to the point that it is no longer considered the same engine? My question is subjective, I think it's an eye-of-beholder scenario, and I'm not really sure except to re-phrase it as, "will these hypothetical cylinder heads, if installed, make the engine not be a Stovebolt 235?"
Whatever the thoughts are, the fact still remains that good OEM cylinder heads are becoming harder and harder to find. I think you're on the right track.
Brad