I am not aware of any specific protocol, nor do I think there needs to be. I think that every academic is somehow responsible for the work in their published papers, and that responsibility includes at least some degree of responsiveness to questions about those papers. But, as usual, some ways of going about this will do better in eliciting responses. For me, I would want to know that the person who is contacting me about my paper (i) actually understands it, at least to a sufficient degree and (ii) that corresponding with them about my work will actually be helpful. This includes some degree of knowledge of what they want to do with the work in question.
[I should say though that I work in parts of pure mathematics in which it would take a truly brilliant mind to find direct industrial applications. So most of my correspondence is from students, other mathematicians, or people who aspire to be one of the above. For people who work in more applied fields, please discount this answer accordingly.]
Should they disclose their affiliations right away? Should the academic ask for it right away?
If you mean "I am asking this question on behalf of my work for Company X": I think so, yes. When correspondents do not identify themselves in this way, I am less likely to respond at all, and if I do, I will usually ask for some identifying information.
Should they say 'I don't want to tell you why I'm asking this for, but can you please tell me so and so'?
Does anyone want to answer a question which is framed in that manner? I would help out a good friend who came to me in this way, but not a stranger.
In general, asking for help from an academic and not being willing to be forthcoming about what it is used for doesn't sound right to me. If you are just prevailing on the goodwill of an academic to publicly explain her work, then have the good grace to say what you need it for. If on the other hand you are actually trying to extract further expertise or information used for some proprietary purpose, then just writing to ask for it doesn't sound appropriate to me. If you want to use someone's professional expertise for some proprietary purpose -- or really any purpose other than just advancing your own knowledge -- then you should invite the professional to enter into a paid consulting relationship, it seems to me. In such a relationship, how much information about the use of the professional's expertise will be provided is something to be negotiated in advance.