We should start by asking ourselves what the heck the difference between a canon and a series is. (Also, what is a project then? So many questions.)
The implicit definition here is quite interesting indeed. Yet the "official" definition from our CI Terms reads as follows:
A list of works by one or more authors portraying a general law, rule or principle.
This leaves much open for interpretation, but interestingly conflicts with multi-author requirement outlined here.
However, I enthusiastically agree with the gist of this definition: a canon is (or should be) less a given narrative sequence (that's what we have series for) but rather an overall mindset / special setting to write in which clashes with the basic premise of articles foreign to this canon.
Now bear with me for a moment on this tangent. There is the concept that fiction is like reality, unless otherwise noted. I think this also holds true for our nebulous "default canon". Whilst we have never outlined it in detail, certain things are core parts of our premise: the Insurgency is in the darker grey spectrum of morality, there are other anomalous GoI around that oppose it, etc. Articles that disregard some of these unspoken base concepts have not just an effect on suspension of disbelief, but canonicity as a whole. (Think A Possible World for one such example: it clearly is not portraying the oh-so-elusive status quo.)
Canons could/would/should be harbours for such outliers with altered premise. To take examples from the other Wiki, I think Broken Masquerade (the Foundation is not secret) or End of Death (everyone suddenly became immortal) are good examples of canons that add their own unique twists to the premise.
However, I often see Canons used as you proposed: as extended series. They usually lack the element of mutual exclusivity. To learn from the other wiki's mistakes, On guard 43 (takes place in Canada) The Gulf (guess where this plays) and AIAD (focused on Foundation AIs) could all co-exist and are suspiciously structured like a Tale series.
As mentioned before, we already have series for that. Those can easily incorporate spin-offs/bonus arcs/inspired works by third-parties without anything catching fire. If we need to distinguish between the central works of the initial author and the "expanded universe" around it, we can just give an own name and category.
That would also help you in your case, as it would be much simpler to open Moving Shadows to contribution. Like having an "Open" Series and a "Closed" Series.
Which leads me to another tangent: so far we only have one "canon" on the wiki, and this is Sectors and Steel. I don't think this thing is really a canon and could be easily downgraded to a particular setting/series (especially given how it currently appears to have only one surviving entry). Equal treatment for all.
So, circling back to your proposal with canon splits, I see a couple of issues. First of all, it's quite a lot of work, and I don't trust that this will be adhered to consequently. Also, it may involve quite a bunch of arguing which could in fact just be used to directly discuss the piece's canonicity.
Then, I wonder what happens if a canon splits (an event that might be very rare)? Do we now have to do the same for each of them, in case either splits again? That would lead to exponential complexity (and administrative overhead).
Lastly, this is a collaborative writing effort. The entire point of this is being able to remix and use as much of the existing work as you want and inspire each other. Think of this like Jazz or Doujinshi culture.
So, this was one wall of text if there ever was one, but I had to get a couple things of my chest. However, I think that the way we'll handle Moving Shadows will be precedent for future cases, so we probably shouldn't just handwave this.
