Aloha Mario, I like to cross the turns also. That reduces the "one turn effect."
I've worked on small counterpoises off and on over the years and took another look at this recently for two 160 meter ground plane builds. The end result was that one antenna ended up with 4 x 40m radials and the other 2 x 40m radials. The radials are about 2 meters above ground. Obviously we had the room for full size radials.
In EZNEC (NEC-2 engine) I see more ground loss using small counterpoises as compared to 1/4 wavelength radials and attribute this to a higher E-field between the counterpoise and the ground. QST recently had an article on a spiral counterpoise that looks attractive if space is really at a premium. A 160 meter spiral counterpoise can be built having a 10 meter diameter. EZNEC shows this to have more ground loss than a larger counterpoise.
I compared an 80 meter spiral counterpoise and FCP counterpoise inverted-L. The spiral shows higher input resistance which I believe indicates greater ground loss. The gain difference is 0.3 dB which agrees.
Spiral, 28 ohms
FCP, 25 ohms
N6LF wrote a series of QEX papers on verticals and elevated counterpoises.
https://www.antennasbyn6lf.com/design_of_radial_ground_systems/Short Radials for Ground Plane Antennas, by N6LFhttps://rudys.typepad.com/ant/files/antenna_vertical_short_radials.pdfFig 4 shows elevated radials at 1' and 10' over ground. For the base-referred GND resistance subtract 36 ohms from the feedpoint resistance.
LF-MF Ground Systems, by N6LF https://rudys.typepad.com/files/chapter-5-.pdf