Those that say that lowering the horizontal tails will do the same thing as the anhedral perhaps can point to the cause of the effect. Also the effect being enhanced with aft CG also point to one thing. Something that causes a small change in horizontal tail load due to yaw), which you have pointed out. Where does it come from?

Flying straight and level at 1 g with the symmetrical wing does produce a downwash that decreases with distance aft of the wing and varies with the vertical position that is being considered. Smaller as you go up from the wing plane.

The horizontal tail has an upload if the airplane is trimmed (wing lift times distance from 1/4 chord to GC = tail lift times distance from 1/4 of tail to CG). With a high mounted horizontal tail location the downwash is pretty small so the tail angle that is required to make the moments about CG equal to zero has a small component due to downwash.

Yawing the airplane disturbs the otherwise reasonably smooth flowfield around the fuselage probably causing a slight decrease in dynamic pressure near the fuselage creating a decrease in lift around that general area of the wing and a decrease in downwash. So with yaw at the location of the higher horizontal tail you have a small local angle of attack change (due to downwash change) and a lower dynamic pressure. The lower dynamic pressure (which decreases the upload on the tail) and the decreased downwash (which increases the upload on the tail) gives a net airplane nose down moment.

Again, as you said an aft CG will make the airplane more responsive to tail load changes due to a decrease in aircraft stability.

If the above is true and if the guys saying that a lower horizontal tail is acting the same as an anhedral tail are right then moving the horizontal tail down or introducing anhedral (effectively lowering the horizontal tail) has apparently moved the tail to an area where the net result of change in tail lift occuring due to yaw is small.

Since the loss in dynamic pressure at the tail is close to the same with either horizontal tail position (for the same yaw angle) the nose down pitching effect, or lack of it, must come from the change in downwash at the new lower horizontal position.

Since at the lowered horizontal tail position the local angle of attack due to downwash is greater (more down) the airplane would require a retrim of the horizontal slightly leading edge up to keep the overall airplane pitching moment the same as before the lowered/anhedraled tail was enabled.

If an assumption is made that the change in downwash due to yaw is close to the same with either high or low horizontal tail position and if the total downwash is greater at the low horizontal tail position then the percentage change in pitching moment due to downwash will be less for the case where the horizontal tail is in the low position.

This gives you an airplane that will respond less in pitch when yawed in level upright flight.

I have probably missed something because it is late and I am getting sleepy but it seems like it should be something similiar to the above.