I read the dialog, and I do not agree with your description of it, as a paradox of GR. The fact is, as I said, the "GR description" is actually not GR at all, it's SR being expressed in noninertial coordinates. The use of the term "gravitational field" is picturesque and suggestive, but nothing that Einstein says about interpreting the twin paradox from the point of view of the traveling twin is in any way dependent on Einstein's theory of gravity. As I have pointed out, the relationship goes the other way around: Einstein's theory of gravity piggy-backs on SR as expressed in non-inertial coordinates.
If you start with ordinary coordinates (x,t), and do a transformation to noninertial accelerated coordinates, then in these new coordinates there are weird effects:
1. An unsupported object will spontaneously accelerate "downward". A force must be exterted to keep a massive object "at rest".
2. For two clocks at rest at different "heights", the one that is higher will run faster (that is, dT/dt is greater, where T is the time on the clock, and t is coordinate time).
These are *not* inferences from GR. They are inferences from *Special Relativity* that are derivable using calculus. You can picturesquely describe an accelerated coordinate system in terms of "gravitational fields" as an explanation for point 1, but that has no physical content. You are just giving a name to the effect in point 1.
To reiterate, I think that you have completely misinterpreted the twin paradox if you believe that is a consistency issue for General Relativity. It is not. The only "General Relativity" that Einstein uses in that dialog is SR + noninertial coordinates, which cannot *POSSIBLY* be inconsistent unless SR is.