I have been following the discussion generated by reactions to Krugman with some interest, as a hagiographer who has dealt with some of the ancients mentioned in the various communications. 

My rhetorical question regarding the discussion is: why 1936 and not 1937?

Indeed, in 1937 the model which formed the basis for SI-LL Keynesianism  appeared in published  form in a number of places, and generated the attempts by Lange (1938), Samuelson (1941), Timlin (1942)  Modigliani (1944) , and Klein (1947) and others to extend, generalize,  and even dynamize it; forging the "neo-classical synthesis" that became the "trained intuition of economists", as Solow put it...SI-LL, in its variant forms, is still taught today, and even forms that basis for more sophisticated macromodels; I do not know of any mainstream undergraduate or graduate text that presents the "model of Keynes"  as related by the Cambridge "followers" (see, e.g. Pasinetti, 1972).

Another important point to consider. It was the SILL Keynesian of Harrod-Hicks-Meade vintage that brought Hansen into the Keynesian "Camp" , and even convinced Hawtrey of the efficacy of demand management (see Young,1987)...

I think we should distinguish between Keynes of the GT and its "policy prescriptions" and SILL " and its policy prescirptions" in a closed economy, as Leijonhifvuud brilliantly did almost four decades ago....the open economy applications of SILL only strengthen the argument (I put aside the issue of whether the AD-AS model can be generated from SILL, as it is problematic, to say the least).

I know that I may be focusing the discussion in a different direction, but I think that my points may still be relevant, at least to some....

Warren Young