I've done some modeling of similar sized loops for portable antennas for 40/60/80m,
typically in a triangle hanging from a single center mast. I actually had a piece of #8 wire cut
for it, and a string of mica capacitors for testing, but haven't found time to try it since
then. Some things I found (though I'll have to go back to my notes for the details):
Wire size makes a difference in the loop dimensions. Larger wire also helped improve
efficiency, though I don't remember just how much.
I fed mine at the bottom, with equal capacitors on either side of the feedpoint for
better balance. I don't know how critical that is.
The feedpoint resistance increases up to 1/2 wavelength, then drops back as the perimeter
approaches 1 wavelength. That means not only is there a 50 ohm point, but also
a 200 ohm point where a 4 : 1 balun would give a good match. For some reason I
seemed to have preferred the higher impedance, but I don't remember if it was due
to wider bandwidth, easier values for multiband operation, or what.
I did develop some designs that worked several bands on the same loop by changing
the matching components at the feedpoint. Banana plugs might be a good switching
method for portable use.
It certainly has good possibilities, and there is a lot of flexibility in the loop shape,
though the optimum dimensions for a good match will vary somewhat with shape.