TL;DR: the question was closed as poorly asked, not as off topic. If it were better asked, we'd know with more clarity whether or not it's on topic. However, non-setting-specific research questions about non-mechanical elements of play are generally going to be a bad fit for this site.
I think this is getting mis-framed. There's no doubt that the question can be answered, nor that it can be relevant to the RPG experience. Arguments based on that line of reasoning aren't going to get anywhere because they aren't what's being challenged.
This is a matter of how wide RPG.SE can cast its net without losing its focus, and it's a discussion which has been going on for some time.
An important thing to remember is that SE is explicitly uninterested in being like other sites. A common evaluative tool for site scope is to ask whether another site would do it equally well or better--in this case, RPG forums. This is why "it's useful" isn't always sufficient to justify on-topic-ness--for example, questions which solicit unending iterative lists are perfectly suited for forums but very poorly suited for SE.
These meta threads seem to be the most useful for seeing the reasoning behind the site's current approach to the site's scope for setting/worldbuilding questions:
Are campaign research questions on topic?
/ Are campaign research questions on topic, part two?
Why did the automobiles and telephones thread close?
What do games like Cthulhu require of Campaign Research Questions?
Are 'fluff' setting related questions relevant?
In particular I draw our attention to this observation about the divergence between the community's theory and the community's practice.
You'll see that it's not cut and dried, but there is a sense that a line must be drawn--otherwise everything becomes on-topic because there are RPGs about everything. Exactly where that line's drawn is an ongoing community-driven conversation. At the moment, the community seems to think science-based non-system-influenced questions about fantasy biology is a little too far afield.
Personally, I feel that if the Animal Planet channel could do a special which satisfactorily answers the question, it's not specific enough to RPGs for this site to field without excessively diluting our focus. (See: mermaids, dragons, etc.)
This question is speculative and framed as an abstract ponderment rather than a practical request for advice about a challenge or problem the querent is facing. This puts up warning flags, because such questions are usually either invitations to discussion or have something very specific in mind which they haven't communicated (so answers are just guessing what problem we're being asked to solve, and the querent is likely to respond "but what about X thing I didn't mention?"). When this kind of question becomes a regular problem an SE site either bans them entirely or imposes stricter limits on them (see system-recommendation). The thing to do is to ask the querent to edit the question so it includes what prompted them to ask it.
The subject material in this case is unaddressed by its RPG context (bowel movements in D&D are purely the subject of third party and homebrew content), and while plausible biology-based answers can be provided, they will always be founded on shaky assumptions which makes them opinion-based rather than objectively accurate answers (for example, the assumption that a creature which looks like two other creatures but has abilities like darkvision shared by neither of them is similar to them in ways other than gross appearance--an assumption proven false multiple times in D&D by the presence of various "mimic" style monsters). The question is not asking about behavioural choices, but only about physical ability, so the setting's social context (and it's useful to note that dnd-3.5e encompasses many diverse settings rather than being itself a setting tag) is irrelevant to answers unless we read more into the question than is actually there--in which case we should, again, be asking the querent for clarification to improve the question instead of answering based on our interpretation of insufficient information.
The best answer for the question as written is "D&D 3.5 doesn't tell us anything about that," followed by some advice about where to go from there based on the answerer's experience with bowel movements in RPGs (there's something I never thought I'd type). See this question and its answer as an example of a question which gives context and an answer which gives guidance rather than speculation. Any further answer must necessarily be re-framing the question, which is fine but should follow the guidelines laid out in this thread--and again, it'd probably be better to ask the querent for clarification.
