As long as you're keeping the aluminum-foil industry profitable (don't forget to save some for a hat to wear on the balcony during these troubled geomagnetic times), you might consider a two-turn loop. That would quadruple the inductance, meaning a smaller tuning capacitance, and higher radiation resistance. Ohmic losses become less critical. Reduced current for a given power level means higher terminal voltages, but at QRP power levels they should be manageable.
There's always a catch though. The current flow in adjacent turns may lead to additional losses:
Unfortunately I've never seen a calculator that attempts to quantify these losses. I guess wide separation between the turns is called for.
EDIT: wide separation won't increase the inductance however. So if the point is to increase the inductance, any loss caused by the proximity effect of adjacent windings must simply be accepted as part of the deal. I have no idea of the amount of this loss, but it might be worth a try (and lead to a more compact antenna). Anyway, I'm not expecting miracles on 40m, though it would be nice to at least radiate some energy there, as it seems active most any time of day here, unlike the higher daytime-only bands.