One factor that's worth noting is that the definition of LGBT+ has expanded to be more inclusive. 20 years ago, an asexual person for instance might not even know that asexuality is a thing, and might have felt that they don't fit in or qualify as an LGBT person. Likewise, for me, I didn't identify as LGBT bc I didn't want to be with a man as a man, and it was only later that I realized being trans was an option, and of course my exposure early on was transphobic stuff like South Park. There's also, like, sissies were/are a thing in BDSM with AMAB people who might be into feminization/crossdressing, but might also be more into women and therefore might not have identified as LGBT before, but now I'm sure some identify as NB/trans.

Having said that, I don't think that sexuality is as set in stone as people think it is. To some extent, saying that people are naturally LGBT or not-LGBT kind of assumes that categories of gender are innate, as well as attraction. But if that were true, then standards of beauty wouldn't change over time.

If we want to dissect this, like, some people are turned on by high heels, and I think it should be obvious that there's no "high heel gene" that determines that. Rather, heels have a certain meaning based on socially constructed associations. Originally, heels were worn by men looking to appear taller and therefore more masculine, but over time they shifted to be associated with women. A person might enjoy wearing heels themselves, or they might enjoy getting stepped on, or any number of other interactions with them, and these interactions have a psychological meaning to the people involved, and may fulfill psychological needs and desires that are more complex than simple attraction. If a person feels pressured in their day to day life, they may seek to hand over control in a sexual context, for example. A person's psychological needs are dependent on their material conditions, meanwhile, the meaning that they assign to particular stimuli are very much influenced by culture. And if we can say that being attracted to heels is determined by social conditioning and the cultural meaning assigned to them, then it doesn't seem like a stretch to say the same about, say, breasts.

So hypothetically, would it be possible to like, raise a kid in a lab and control whether they're attracted to men or women? I think it'd be possible to tip the scales, but the human mind is very complex and I doubt that it's possible to predict things to that extent. If you try to make them be attracted to one thing, they might discover another thing and find it new and exciting instead. Like, conversion therapy and stuff can just make it so that being straight is associated with doing what your parents and society want you to do, while being gay is associated with defying that and doing what you want. So it's not something that can be changed so bluntly. But I think it stands to reason that generally speaking, social stigma is going to lead to fewer people identifying a certain way, and that's even if it's an anonymous survey, because people are going to be less likely to investigate feelings or be open to new experiences, and to admit things to themselves.