Factors I'm using
- Preferably, it should be about some subject that they have already learned
- If possible, there should be only two big theories (and maybe a few alternatives with minority support), with the debate being mostly about which of them fits the available data better
- It shouldn't be too politicized (cuts off most of the "was X real?" options) and/or too focused on emotions (cuts off most of the "why X did Y?" options)
- Nevertheless, both theories should have fairly wide level of support; it shouldn't just be "one guy has a theory that disagrees with everyone else", so e.g. Marco Polo is out
With that in mind, here's my current proposals; I might add some more if I manage to think of any.
(I actually only had the first one [now second] until I almost finished writing it down and then suddenly thought of another [now third].)
EDIT: I've asked my brother (15 years old, passionate about history) and got an answer that I liked so much I'm putting it first. (I'm keeping the other two to avoid confusing the upvoters too much.)
Death of Alexander the Great
Alexander III of Macedon, aka Alexander the Great, conquered most of the then-known world before unexpectedly dying in 323 BC at age 32 (a month short of 33) in Babylon after about two weeks of illness.
People, especially rich people, didn't particularly often randomly die of natural causes in their 30s back then any more than they do now.
So did he die of an infection? Was he poisoned? Did he perhaps get a food poisoning or something? If it was a murder, then why, and how?
Historians debate these questions to this day.
Possible bonus inclusion: where he was buried, and what happened to his tomb? Its location is currently unknown, and legends abound.
- Alexander the Great was probably the most important person in the entirety of ancient history, and his sudden death was arguably the most important event in the entirety of ancient history
- The political implications of his death, to the extent that there are any, are for the most part about it having occurred at all, rather than the specific cause, so there aren't a lot of politically-motivated theories trying to prove a particular scenario
- Kids love murder mysteries 
- To a large extent, the true answer is unknowable; there's just not that much evidence, especially regarding possible reasons to kill him
- 12-year-old children might be too young for an extended discussion about sickness and poison symptoms
The other two options (which I came up with by myself) are listed below.
Chronology of Mesopotamia: middle vs. short (and the other alternatives)
TL/DR: when those kids learned about Mesopotamian civilizations, they probably received some dates associated with them. Which dates? Were those dates correct? And how do we know?
This might be a bit too complicated (and/or too vague) for a "major historical event", but this is certainly a debate that historians are still uncertain of to this day, with evidence on both sides, despite sources.
The chronology of Mesopotamia is, to a large extent, floating; there's usually (though not always) enough evidence to say that event A happened X years after event B (there's a lot of those cuneiform tablets, and we're constantly finding even more), but there's not necessarily enough evidence to say that any particular event C (in the earlier areas, anyway) happened Y years before the common era.
There's a bunch of astronomical evidence (most notably the Venus tablets of Ammisaduqa) allowing historians to fix the options to only a few possibilities, and a bunch more extra evidence (e.g. radiocarbon, dendrochronology, inter-regional synchronisms) that can be interpreted to favor a particular side (though which side it favors can itself be disputed).
There are similar considerations in Egyptian chronology, but the situation for that is quite messed up (not helped in the least by the main source for it being in absolute tatters), whereas for Mesopotamia historians, for the most part, have a very good idea that is down to only a few possibilities.
- Limited to only two major options (plus some minority candidates, still not many - maybe four or five - outside the extreme fringe)
- Essentially civil; approximately nobody cares all that much about when exactly the events happened (except the fringe, which for the most part doesn't accept any of the mainstream options anyway), so there isn't much of a heated dispute
- The general subject of "how do we know when this happened", and even more generally "how do we know what we know about history", is very important to consider, especially if anyone in the class is thinking about becoming a historian
- Approximately nobody cares all that much about when exactly the events happened (and many of the people who do care don't like any of the mainstream options), so the dispute is of fairly low importance
- The evidence for or against the different chronologies, or possibly even the chronologies themselves, might be a little too arcane for 12-year-olds to understand
- There isn't really anything cool riding on it one way or another; it's just so much numbers
Was the Indus script (and/or the Easter Island script) a full writing system, or just some pictograms that weren't intended to be read as a language? Or even something in between?
Ever since the seals at Mohenjo-Daro (and/or the rongo-rongo tablets) were discovered, historians had debated whether they were, in fact, a real writing system for a real language, or just some generic pictograms (something like logos, desktop icons, or emoji, to take comparisons intelligible to 12-year-olds).
There's a bunch of evidence on both sides, but not all that much for either, and the civilizations in question are sufficiently far gone (and were sufficiently unconnected with other civilizations that historians do know more about) that there's not much non-internal evidence to be gained (outside a particularly lucky break).
So historians basically have to deal with what little evidence they do have - which results in some unexpected considerations.
Possible bonus inclusions: Phaistos disk, Voynich manuscript, quipu.
- Not just a bunch of numbers but something with actually interesting relevance
- The general subject of undeciphered scripts - in some cases (e.g. Linear A) we know that something is a script but don't know (much about) how it is supposed to be read; in some other cases (e.g. Etruscan) we know how it is read but not necessarily what it means
- You can also talk with your students about whether emoji count as a language, and/or ask them to imagine a bunch of signs written in emoji found by a future civilization
- The question invites a lot of kooky opinions (particularly of the spiritual variety), and there are so many of those that they start overshadowing the genuine opinions
- In the case of Indus script the question is also highly politicized, because the authorities of India really want people to believe that the Indus Valley civilization is a direct ancestor of later Indian culture
- The evidence favoring one interpretation over another probably won't be easy to explain to 12-year-olds (not that it's particularly easier to explain to serious historians)
- If your students hadn't heard of the Indus Valley civilization they might be lost at the very start (though they probably have at least heard of the huge heads from Easter Island)