The most problematic part, and it is a crucial part of Piaget as I understand him, is that the constructive synthesis comes about through the "functional overlaps"--that is through the way the schemes are functional in regard to our knowing the external world.  This idea is supposed to ensure that the new structures retain and improve upon the the earlier ones in regard to their functionality for knowing the world.  (That is to say, if the new structures were generated from mechanisms of coordination that did not include guidance form the external world, the mechanism for producing new structures would not necessarily produce structures that were more functional for knowing the external world than the earlier one, and one would need to account for this in some other way).   I think that this puts a big wrench in the relationship Piaget tried to draw between the intellectual's development through qualitatively different structures, on one hand, and the mechanisms that account for the changing conceptualizations we develop in scientific activity.  In the latter case surely no method intrinsic to the generation of new systems of  ideas guarantees their greater functionality than previously existing systems of  ideas.