It is a magnetic loop design. Note the "Let's look at the ideal coupling method:" here:
https://www.w8ji.com/magnetic_receiving_loops.htmAn often-cited claim for a shielded loop is the design somehow blocking electrical fields to optimize reception of magnetic fields. Outside of maybe some slight near-field effects, a shielded loop does not really shield much of anything.
Instead the shielded design is more or less to help balance loop symmetry to potentially equalize differential currents and limit common-mode currents. Therein is why a "shielded" loop might improve reception if CMC is present on the feedline. If not, then there is not much, if any difference versus just a simple wire loop.
These days, I might would make some changes. I probably would use a mix 73 binocular ferrite with a 1:1 winding ratio at the feedpoint. Using 4 turns towards the loop and 4 turns towards the feedline should suffice in my experience. The result is an isolation transformer. You probably could just float the loop shield in such a config.
BTW, another often-cited LoG design consideration is software modelling might suggest a couple or so dB in directionality if feeding the loop at a corner instead of a midpoint along a side. Being realistic I kind of doubt any noticeable real-world difference.
Another popular alternative for a coaxial loop would be the NCPL design, which is basically a mobius loop. It might help a little with low-frequency performance for a typical-sized loop-on-ground antenna.
https://swling.com/blog/2020/04/diy-how-to-build-a-noise-cancelling-passive-loop-ncpl-antenna/https://www.hamradio.me/antennas/airspy-youloop-lf-mf-hf-mobius-receive-antenna.html
All the above said, unless you have the coax available for experimenting, I would suggest first trying a simple wire loop-on-ground. 60' to 150' of whatever wire should suffice for MW to mid-HF. You might need a preamp towards upper-HF. Use an 1:1 balun or an isolation transformer at the feepoint, plus perhaps similar back near the receiver.
Case in point, my old 148' "shielded" LoG ended up buried under leaves, pine needles, and shifting soil/sand. A typical LoG is already well into negative dBs, and going underground can make the losses even worse. Mine was old enough to not bother with it anymore, so I eventually just deployed another LoG at a different spot. My current LoG is 148' of copper pet fence boundary wire with a LDG 1:1 balun at the feedpoint and a mix 73 isolation transformer near the receiver. The simple wire LoG seems to suffice for my purposes at the moment. YMMV of course.