I would think that the inverted clay pot would reduce the wicking, which exacerbates the dry top.
In all cases, we are playing a balancing act between the evaporation rate from the top of the medium and the ability of the medium to replace the moisture through wicking.
Let's use an example in which a pellet is capable of transferring 90% of its moisture to the pellets around it (I really have no idea what the actual percentage is, so don't quote me on this).
- The first transfer out of the reservoir, and the pellets that are 100% saturated wet the next ones to 90%. The layer above that 81%, then 73%, 66%, 59% 53%, and 48% at our theoretical top surface. (Add more layers and it gets lower - hence the difference between tall and squat pots.)
- If evaporation - due to temperature, air movement, and ambient humidity - is fast, the 48% wet top surface will lose most of its moisture, the layer below that (because it is somewhat "protected" by the layer above) will lose a bit less, and so-on down into the pot. As those layers have reduced moisture, they transfer less upward.
- Eventually, a balance is achieved, but if any of the ambient conditions change, a new balance has to be established.
Going back to that first bulleted paragraph, now let's put a pot in the reservoir. While a clay pot does wick water, based purely on surface area, it will not do nearly as much as the pellets it displaced, so your transfer upward is reduced right from the start.