Absent any other external forces, liquids (even non-uniform ones like the ocean) will form an isograv, where every part of their surface is at an equal gravitat

Absent any other external forces, liquids (even non-uniform ones like the ocean) will form an isograv, where every part of their surface is at an equal gravitational potential. This is a natural consequence of how pressure works. So you expect this kind of deviation from gravitational anomaly. And while other forces can have a significant impact, there's not a lot else which can cause a persistent change in the average of this size. Dynamic effects will just to make things oscillate around the average but not chang the average much if at all, and even an extremely heavy prevailing wind (rare at the surface) won't be enough to cause much of a change in sea level. (ok, there is one really big thing left off of this which is important with respect to what 'average sea level' means: the largest other persistent effect is the centrifugal effects from the spinning of the earth, which results in the bulging at the equator (both for the rock itself, which at planet scales and timelines is approximately liquid, and for the oceans). This is way larger: the oceans are actually about 7km higher near the equator than you would expect if the earth was a sphere. The average they are comparing to here is coming from measuring sea level with respect to the shape it would be if the earth were just a ball of liquid with the same mass and spin. There's a bunch of different approximations which get used here and the details get really messy, but the next step after taking into account the centrifugal effect is basically various levels of approximation of the distribution of gravity: https://en.wikipedia.org/wiki/Geoid https://en.wikipedia.org/wiki/Geoid)