If you consider groundwater to be an infinite reservoir, and relative to a pond it will be, the pressure will equilibrate whether pond is 4' deep or 8' deep. There will be greater pressure down at 8', but the hydrostatic pressure is also greater and will balance. If the small tank's bottom moves down 2', the large tank will fill in to the same level. If two tanks have 3' of water in them, the level won't change if you pipe across 1' down or 3' down.

If the connection to groundwater is in the bottom of the pond, it will keep dropping as levels ebb. If the connection is 2' down, then theoretically it will quit leaking (badly) once you lose those 2' of water.

If you are concerned about transients, the filling and emptying process, if you open up more "conduction paths" to the reservoir, you will have greater flux in and out. Since the proposed work is on a filled pond and I didn't see mention of draining and keeping pumped out, compacting layers isn't a means to seal off the connection to groundwater. If that's true, any digging can only open up more routes in/out.

We don't know if the groundwater enters at 2' depth or at the bottom do we? If you know where your opening is, I would say dig in the other direction. If it's on the bottom, then widening and cutting soil away at 2' depth would seem least likely to increase leakage rates. If you have a ring at 2' depth, then going deeper would seem to least likely. If the whole thing is porous, then you'll just get more flow in and out during transients.

Just opinions from a guy who does fluid flow when not at the pond. Best of luck with the project. We're all rooting for you.