Thanks for putting up those images.
One is clearly better than the other but it is a bit hard to work it out on a phone.
It’s this borne out by the graphs you provided?
Noting that the cookie can antenna is less than a quarter wave length from the edge of the ground plane, it is within half a wavelength of a resonant element, it is magnetic based pierce of crap widely known as the poorest ADSB antenna around and with completely unknown impedance yet it is no t worse than a pcb dipole.
I suggest there is really something amiss with the dipole or the test process itself is questionable.
Here is some interesting Research into ground plane antenna .
As yet i haven’t found any convincing research that a PCB dipole antenna behaves the same as a true coaxial dipole.
The coaxial dipole was the vhf and uhf antenna of choice for vertical mount from late 1960s through to the 1990s when computer designed multi element arrays in a raydome proved cheap enough to manufacture…