Sorry Tim, I beg to differ....

True, the tank circuit frequency gets dragged around by "hand"
capacitance, but the place where ths effect is felt is almost
invariably the capacitance between ANTENNA, CHASSIS, BODY and GROUND
rather than direct hand-to-LC.

Better grounding is nearly useless, unless you bolt the chassis to a
very good, high capacitance ground - e.g. you bolt it to a steel bridge
on a ship. But even then, your hand and body may introduce a change
in antenna-to-ground capacitance - back to square one!

An extremely loose coupling will reduce this effect, but it's very
hard to pull off while keeping an efficient energy transfer.

It's better if you use an active buffer stage. Nowadays you rarely see
solid state regens without a buffer, and the reason is not signal
levels - it's that capacitance decoupling through an active device can
be very effective, making both frequency and regeneration more
controllable.

Another trick, less practical, is to move the antenna far away, and
feed it via coax. In this case, grounding the coax somewhere along the
run (at zero impedance if you route the coax by a good RF ground point)
is very effective at stabilizing antenna capacitance to ground. In
this case you will indeed see almost exclusively the effect of
hand-to-LC, which won't amount to much and can indeed be fully killed
by a shielded enclosure.