Moho

That's not what I would have expected. The ESA's Goce satellite is mapping the depth of the Moho boundary", the transition between crust and mantle. Some of the features seem to agree well with what we know: mid-ocean ridges and the Himalayas, for example. Also the Andes mountains: as they are higher above the surface, so also the depth below is deeper. But not the Rockies; the boundary there is normal. The deepest North American Moho boundaries are under the Great Plains instead. I can't quite figure out why the continental plates should ride differently in South America and North--I thought both were subducting the ocean plates. And possibly the glaciers planed off elevations in the Plains.

In any event, the effects in the Himalayas are quite dramatic: Everest is about 9km high, but the Moho boundary is pressed 30km deeper below the Himalayas than below the surrounding land; and similarly under the Andes.

You can't expect perfect equilibrium--for the weight of rock above to equal the weight of rock pushed down. Spalling and erosion will remove a lot of the excess above, and on a timescale much faster than continental drift. Still, that's a lot of rock "under water".

Of course there may be some biases in their reconstruction; they can't measure the depth directly, and there could be some error in the correction terms they use to account for the continental rock.

Something else to keep an eye on.