This is how I read the question, which is admittedly a bit more concrete than the question text:
What are huge experiments which are going to run in the future and are currently under development (which means that at least part of the funding is already secured and people start building something). By "huge experiments" I mean experiments of the LHC,LIGO type, which means that a) thousands of people of all continents are involved and spend or will spend a majority of their time on the project and b) the experiment really is one big experiment (such as one telescope, one detector, etc.) and not a very tightly connected net of smaller experiments.
As such, it is a big-list question and, as you state, in principle off-topic. I don't think this is a reason to "close as off-topic", because it's just a general "big lists are off-topic". We should have a look at why big lists are off-topic. The best (?) reasons I could find are summarised by DavidZ: Why exactly are big-list questions discouraged? Shouldn't the FAQ say this prominently?
List questions are junk food for the site, in a way; they tend to be popular and attract a lot of attention, but they don't provide a basis for judging the correctness of answers (so the voting system is wasted on them) and they never really end, thus drawing attention away from other, better questions.
I would argue that these reasons are invalid for this particular question and that this particular question fits into the framework. Here are my reasons:
The question (as stated above) is precise, i.e. there is an objective way of judging the correctness of the answer: Is the experiment truly huge? Is it a physics experiment trying to explore physical concepts (on this ground, ITER might not really be a physics experiment) Is it past the initial exploration phase and under development? Is the experiment anticipated by parts of the physics community? This means that it is not overly opinion-based either.
The question is sufficiently focused. Indeed, I would argue that it is not too broad per se: the number of experiments that fits the bill won't be too huge. I bet there are less than twenty experiments.
The Stack Exchange framework does allow for a number of good answers and can, in principle, support small lists (I'm not so sure about lists like on mathoverflow with dozens of pages) and it is particularly good for questions where maybe not one person has the complete answer, but a complete picture emerges from several good answers.
While the question might not really "end", activity is very restricted. Only every few years will there be new experiments that fit the bill. Also, by the nature of the question, even without additions it will stay up-to-date for several years, if not a decade. For me, this is stable enough (who knows what happens with Stack Exchange in ten years?).
The question is interesting to researchers and enthusiasts alike. Hardly anyone knows all of these experiments, but many like to know whats going on outside their fields at least a bit.
There are two points that might be problematic:
In a sense, this gives a lot of reputation for easy answers, because of the popularity of the question. To counteract this, one might impose a similar rule as on mathoverflow, where all answers are immediately community-wiki. Then again, almost everybody gets their reputation from easy answers, so maybe this is not a real problem.
Clearly, the question garners a lot of attention. Maybe it draws away attention from better questions. I do not have a good counterargument other than that I feel this is true for almost all hot network questions, whether they are on-topic or not. Often, the physical content is even less than one can learn from the answers to this question.
In summary: I think that this is one of the very rare instances where such a list-based question should actually be on-topic for this site and I argued that the reasoning behind the close-policies allow for such a question to be off-topic.