I think you're spot on with all of this. The only way to do good molecular/cell biological benchwork is with access to a bona-fide lab. I have a suspicion that I could ferret out most of the kids who didn't do the bio work, but I suppose there could be a few who know the nuances of the work really well (or they can make up plausible lies and state them with confidence).
It might be better to restrict these competitions to "engineering". You can do some pretty good engineering/programming even as a youngster.
>it's much harder for judges to sniff out people who didn't do original work in the biological sciences category
This applies to biological research all this way up the chain, including research done at top Universities, and by "top men".
Perhaps we should only fund the women then.
I went to the state science fair years ago in 8th grade, and nearly every single person I met there had a parent who was heavily involved in their project.
You could tell what each kids parent did from a living just by looking at the posters. The CS project kids had parents who were programmers, the kids with engineering projects had parents who were engineers etc...
Here's my perspective on the whole thing. Take it as you will.
At the regional science fairs (where everyone starts out), there are a lot of people there who are only doing it because they have to. Usually, their high school requires it. A lot of these projects are shit, and a lot of them also clearly stole the idea or got help from their parents. All of these get eliminated at the regional level.
Then, we have the state level fairs.
When it comes to engineering projects (mechanical, electrical, or computer), almost 100% of the projects are actually done by the students themselves. It's very easy for engineering judges to tell when someone's bullshitting or didn't actually put in the work, so almost none of the bullshitters make it to the state level. Most of these people don't have any immediate family members who could help them, even if they wanted help.
When it comes to biological sciences (medicine especially) the situation is much different. Almost everyone in this category knows their shit, but it's a lot harder to say if they actually did the work themselves or not. A lot of the projects absolutely require that the students work in industrial lab environments. It's impossible for a student doing MRSA research to work in a lab below BSL-2, for example. A lot of these students get lab access through their parents (who may be doctors or researchers or something), so it's very likely they are working alongside their parent, and it's possible that they are working with the help of their parent. A huge portion of students doing this kind of research do have an immediate family member who can (and, with some likelihood, does) help them.
Because it's much harder for judges to sniff out people who didn't do original work in the biological sciences category, you get a lot more bullshitters in this category making it all the way to the top science fairs. They're still usually experts in whatever they're researching, but they didn't necessarily put in the thought or effort themselves.
When I was at Intel's international fair, the guy who won the top prize claimed to do some sort of cancer-detection research. It was pretty impressive stuff, but after the science fair a whole bunch of incriminating stuff came out. IIRC, there was some situation where his parent had a position of power at some major research university, and he essentially plagiarized the research of some researcher there with minor changes, but it was soon enough after the actual research that no one caught it.
I can't speak as much to the other categories (sociology, mathematics, etc.). I didn't have a lot of experience with those.