Your answer, in part, states that it is just a subset of the desires that exist to which a moral-ought is relative:This is correct. It makes no sense to consider the desires that are not affected. However it is also important to consider desires that could be affected and not to arbitrarily exclude them prior to the analysis.
Now desires can be directly and indirectly affected. A desire has conditions of fulfilment such that these conditions are met when the proposition(s) that the desire contains are true in some states of the affairs. Some desires might have different conditions of fulfilment that are either fulfilled or thwarted in those same states of affairs. Those are the directly affected desires.
By contrast, other desires are only indirectly affected by such states of affairs. That is the state of affairs and hence the desire that brought it about, are only means or intermediate to other states of affairs that are the targets of those other desires. They are affected, as such states of affairs brought about by the desire under evaluation can help or hinder the realisation of their states of affairs, such helping or hindering being indirect.
Much of the internal critique within desirism is as to what the directly and indirectly affected desires are, particularly indirect desires. That is, accepting the desirist framework, there can still be dispute as to what the affected desires are. The phrase “all desires that exist” serves to ensure that none are excluded on a priori unsound and invalid grounds.
Clearly, then, this is not "all desires that exist". A moral-ought is relative to a subset of all desires that (possibly) exist given your qualifications above. I think this is fine, though. I think the theory still stands. But this "all desires that exist" terminology needs to be clarified to include the qualifications you've pointed out here.I also highlighted in the original post the other internal/external usages of “all desires that exist”. The fact the only some desires are internal to the agent(s) under evaluations does not mean other desires external to them must be excluded. The primary purpose of morality is to help install and promote some desires that the agent lacks, and remove and discourage some other desires that the agents have.
Bearing this caveat over internal/external desires in mind, one could talk about “all affected desires that exist” where appropriate?
I will deal with the other qualification hinted at in the above quotes when I reply to Kip’s final point – over irrational and rational justifications in a future post.