I will bring this to the Tilt-up Concrete Association when we meet with them at World of Concrete in February.
The reinforcing industry obviously needs to clarify what is required and why. When it comes to how much effect there is from "materials deleterious to bond", it seems that we are talking in terms of fractions rather than orders of magnitude. There is a good bit of research going on right now into straight and hooked bar development length, and we expect results on these projects within the next year.
As far as development of deformed bars goes, only about 15-30% or bond strength is generally considered to rely on surface bond between steel and concrete. One of the previously-linked articles noted that "initial slip" was affected. This could very well be the critical part of the equation, since once slip starts, the only resistance remaining is provided by resistance to splitting (that is, concrete tension strength and confining reinforcement and geometry.) Splitting failures of unconfined splices can be brittle and sudden. In the case of a thin wall reinforced in one plane only, reduction of adhesion between bars and concrete could decrease the force required in the bar to induce splitting along the plane of reinforcement.