I'd be concerned about the resistance of the connections to the foil. Most of what I've read wants very good electrical conections with a small transmitting loop.
I agree. The loop design presented above is sub-optimal in that regard, using mechanical pressure connections between the loop's foil conductor and the variable capacitor.
The thing though that I find infinitely cool about magnetic loops is that if you use the loop conductor as the capacitor, then you don't need any electrical connections at all to the loop conductor! (The coupling loop also needs no electrical connection.)
To use the loop conductor itself as a capacitor you just keep the ends of the loop conductor in proximity to each other forming a sort of large gimmick capacitor. I've tried this and it works with 50mm-wide copper tape. Others have inserted copper tubing into itself to form a trombone capacitor - again, with no solder or mechanical connections.
However, once we start to use the crumpled foil not only as the loop inductor but also as the loop capacitor, then we need to worry about the RF losses introduced by the irregular (and large) capacitor surface area - there was talk in another thread about "environmental" losses if capacitor plates are too large, and larger plates also means more stray inductance in what should be a capacitor. Then there's loop balance to worry about as well... still, all in all, if a single piece of foil can make a reasonably efficient small transmitting loop, that would be really, really neat in my book. Consider the usual alternative: silver-soldered copper pipe and a vacuum variable capacitor - orders of magnitude more expensive and troublesome.