You should call that the Crime Scene Interceptor, CSI Antenna.
If we assume the cap is fully meshed (365 pF) at 7 MHz, the loop inductance is about 1.4 uH. Reactance at 7 MHz is 62 ohms. At 21 MHz the reactance is 186 ohms. This reactance variation will be reflected into the coupling loop, so you're never going to have a perfect match to 50 ohms.
Based on the Hoop Loop experience, I wonder if you could put the feed line right across the loop, use flat speaker wire for the feeder, and put the cap at the receiver. This would be for listening, of course, where a 1:1 SWR isn't required. That will save you running dc out to varactors.
The Hoop Loop on MW needs a matching transformer because of its low inductance relative to wavelength. I had to transform that inductance up to something much higher, that the radio's internal variable cap could tune. With this setup I don't think you have to do that, because the 365-pF cap can tune the antenna directly.
If you can rustle up some speaker wire at the local radio store, I'd be very interested in knowing how that works. There's room here for discussion related to feeder capacitance and the effects of connecting a balanced antenna to an unbalanced port (you're doing that now, in any event), but this scheme may work and save you some trouble.