As explained in this Weekly Halacha column by R' Doniel Neustadt, the problem with steam is that the steam from the second dish will rise up, absorb flavor that the first dish had deposited into the oven walls, and then condense and fall back into the second dish, thus adding flavor from the first dish to the second dish. R' Neustadt records a disagreement between the authorities regarding whether the steam making deposits to and withdrawals from the oven walls is plausible in the first place:
According to She’alas Ya’avets, et al., steam dries up and evaporates in a hot oven before it reaches the walls, so the steam concern does not apply to hot ovens, and (as long as they're kept clean), they can be used serially for meat and milk dishes.
According to Igros Moshe, et al., we can't be so sure that the steam evaporates before interacting with the walls, so this concern does apply. However, Igros Moshe qualifies this concern by saying that if the heating element of the oven is on top, then we can be sure that the steam will disappear before contacting the walls of the oven.
I'm not sure how to understand this concept of steam "drying up" and "evaporating," since steam is the product of evaporation. If I'm not mistaken, for the steam (and its flavor cargo) to disappear without interacting with the oven walls, it would have to be physically vented. Perhaps that mechanism is, indeed part of the consideration.
In any case, it's clear that the authorities take into account the possibility that the heat of the oven will prevent the steam from reaching the walls in the first place. Therefore, perhaps this ruling about dishes that come out dry from the Star-K extends that concept to the condensation and return to food process using two additional mitigating factors:
For steam to condense and fall back into food, it would generally have to cool down, which is less probable in an oven that's hot enough to make it rise in the first place.
If the dish starts off dry, then we assume that it's not emitting [significant] steam to begin with, so the steam process won't affect it. If it starts off wet and ends dry, then presumably, any steam that it emits is disappearing above, since it has experienced a significant loss of liquid. So, we can assume that the steam didn't get a chance to return to the dish bearing foreign flavor from the oven walls.
I'm still waiting for a response to my email to the Star-K. I will update this answer when I receive it. In addition, this answer could be improved by looking up the primary sources, checking R' Neustadt's and my interpretations, and editing in links and quotations or direct paraphrases.