Russia: 50 units (plausible)
China: 250 units (plausible)
Rest of the World: 500 units (speculative)
VERY speculative. To the point that even attempting to put a number on it is a bit facile. A truly honest assessment should realistically read "Rest of the World: we don't have a freaking clue?"
With uncertainty of that magnitude inherent in the production estimates, deriving the profit shares from it to split the project between the two partners is rather disingenuous. I'd resolutely object to that method too - find me a criterion that you can actually rely on for making a decision of this kind of importance!
Even if underwritten by the government! There will inevitable some risk remain. As we have seen many times when government have taken on big projects the works over.
Thing is, we don't need to look at big public projects the world over, most of which have at best a tenuous relevance as precedents for the CR929. Why try to gain insights on a civilian aerospace programme underwritten by the Chinese government from hydroelectic dams in Brazil or rail links in Switzerland? We already have a much more specific and pertinent example to refer to: the ARJ21, another civilian aerospace programme underwritten by the Chinese government.
That said, ok the risk in this case is some what reduced due to the potential "captive market" But it's not a given. If all goes wrong, ie the plane never took off (pun intended).
"Somewhat"? In most other countries, the ARJ21 met the definition of a project that "didn't take off" several times over and would in all probability never even have made it as far as first flight. Yet China pressed on, indicating that civilian aerospace is such a high priority that no obstacle is too great.
While perhaps not outright zero, based on this precedent the risk of the CR929 being allowed to fail is so minuscule as to be not worth considering.
On the other hand, turning your argument on its head. If a guarantee market is a risk-free undertaking, then the Russian can simply go ahead on its own without having to share their profit with someone else.
If the Russian home market was the same size as China's you might have a point, but it isn't. From Russian domestic airlines alone you can't get enough volume to make a profit. Apples and oranges.
The Russian are not exposed to any more risk than if they had gone out on their own. In fact the opposite is true. The super duper wings and engines they have is useless as they have hot nothing to attached to.
It's not like the Russians have never built the rest of an airliner (including widebodies) before. In fact, to keep their hand in they are (nonsensically in my opinion) running a minor update of the Il-96 in parallel. A ripe target for cancellation and replacement with a clean-sheet, fully-Russian design if the CR929 falls apart.
It's an obscure fact, but Russia has even done considerable work on CFRP fuselages, to the extent of manufacturing a demonstrator fuselage section for the Il-114 turboprop in the 1990s. Unlike the CR929 which uses the Airbus panel approach this was a wound barrel like the 787, but with integrated geodesic stiffening structure. Both philosophies perform about equally, panels allow you to locally vary skin thickness while winding saves you a lot of fasteners. The Russian solution could have saved even more fasteners than the Boeing approach though, so it's an interesting wildcard.
Your points about the Russian would not benefit from economy of scale is false. Because you are assuming Russians have to make those components themself and for a smaller number of units.
I did not say there would be no benefit, but the difference is substantial. For example, final assembly (if there is a separate Russian line) would in fact see no benefit whatsoever.
What's to stop them from making it for both nations? Or sourcing it from China at a cheaper cost? Both ways to achieve economy of scales.
You mean everybody just builds their part of the plane and all final assembly lines deliver to all markets based on availability? That's very similar to the way Airbus was organized and it is a successful and fair model. Each partner participates equally in the project's fortunes and the risk is shared evenly. If sales are 20% above prediction, production volumes go up 20% for everyone, if they are 50% below, volumes drop 50% for everyone.
It's the antithesis of the proposed deal where Russia carries a disproportionate part of the risk from uncertainty and market volatility*.
Further, the Russians got the captive market of their own country, plus all the rest of the world which was available before. So it's a win-win co-operation. It's only an issue now, because the Russians wanted more.
See above - there is a non-trivial chance that international sales will not materialize so the Russians can't just ignore that contingency their planning.
* Volatility is another aspect where a captive market works greatly in your favour that we hadn't even mentioned yet. While Airbus and Boeing ground to a halt under CoViD-19, COMAC kept delivering ARJ21s more or less according to schedule. If you believe that is because the ARJ21 is such a superb little plane, I have a bridge in to sell you.