NB, in the attached example a Top Hat can be of benefit in raising the current distribution higher up on the radiator, without loss in gain, and maybe a 1* degree increase in maximum angle. Plus of course the issues you make above.
The 1* degree increase may be as a result of this model being only 18' above Earth. If the antennas were higher, this results might not happen.
Theory for the 5/8 wave radiator over the 1/2 wave includes the idea of this increase in the current distribution over a 1/2 wave, so IMO it should hold true in the case of top hats as well. Check out the image below of the side by side Current Reports for wire #3, the radiator. Here we have 20 segments listed for each. The currents are listed upside down in these reports, so the bottom numbers are at the top of the antenna.
Notice from segment #12 up, the maximum current points for both antennas, the antenna with the top hat has the highest currents per segment, and the difference increases nicely all the way up to the tip. More current more higher = more RF higher up.
Again, the arguments are that a full length radiator is always better and in every way, but I say the net differences are probably a toss-up...which is likely undetectable just using our radios.
[ATTACH]6411[/ATTACH]
IMO, I think some manufactures use these small top hats based on the idea that the larger tip might help discourage lightning better than a small tip. This idea is worth considering also, and for sure if we give up little to nothing.
