Per UL 486A-B, a connector for a range of conductor (say #4 through #3/0) will be tested with the largest conductor....#3/0.
But not only that, it will be at 310 Amps for the static heating test, and 345 Amps for the 75 deg rated connector Current-cycling test....
For the static heating test, the allowable temperature rise is 50C, so that may be usefully informative (as far as the physics) as to the thermal behavior at lower currents. Of course you'll have to figure out how to adjust for smaller conductor sizes; the contact resistance may increase with smaller conductors, so it is not enough just to compare currents and ignore the conductor size.
For the current cycling test, the allowable temperature is 125C, so I don't think that will be helpful in arguing that some other practical conditions won't exceed 75C.
Regardless, to the extent that UL 486A-B requires testing at currents in excess of the Table 310.16 values (which it incorporates as the "Assigned maximum ampere rating" column in Table 7), that is presumably to provide a margin of error, and does not constitute permission or even a reason to use ampacities in excess of the Table 310.16 values.
But sure, if I use a #1/0 conductor with a 200A load in a 40degF refrigerated warehouse on that #4-#3/0, 75deg connector, I'm totally going to exceed the temperature rating of the connector because the conductor is smaller than #3/0....I think not.
To give this example the benefit of the doubt, let's say the load is interlocked with a thermostat so that it will never operate at temperatures above 40F (5C). So I infer you'd like to take credit for the ambient temperature being 25C below the 30C design temperature in Table 310.16. A 75C ampacity from that Table is based on a 45C maximum temperature rise over 30C ambient. So with a 70C maximum temperature rise over a 5C ambient, the ampacity would increase by a factor of sqrt(70/45) = 1.248.
The 75C ampacity of 1/0 Cu is 150A per Table 310.16. With the 1.248 temperature correction, if applicable, we would get a 75C ampacity of 187A. So even if this is an allowable method, with the numbers you have chosen, 200A is too much. But we could massage some of the numbers to make it work, e.g. use a 2/0 Cu conductor (75C table ampacity 175A), or lower the load to 185A, or lower the temperature further.
I agree that this method make sense physics-wise. But it is not permitted under 110.14(C) or the AALZ guide-info. That is 100% explicit for the case of AALZ "Distribution and Control Equipment Terminations," which says to use the Table 310.16 values without correction. It's less explicit but still true for 110.14(C) and AALZ "Appliance and Utilization Equipment Terminations."
It is not so uncommon that the regulations do not cover all usage cases with the full allowances that would make sense on the basis of the physics. Simplifications are made to permit ease of administration and understanding by the general use. I have encountered this with multiple PIs to expand the allowances to 705.12(B). In the last round the CMP's response to my proposals was to paraphrase "you're probably right on the physics, but we're not going to further complicate 705.12(B). You can do what you want under 705.12(B)(6)'s "under engineering supervision" provision."
If there is a comparable "under engineering supervision" provision applicable to your example and 110.14(C), then as an engineer you could reasonably stamp a design for the use of 2/0 Cu conductors for a 200A load in your example, or 1/0 Cu for a 185A load. Otherwise, you are out of luck and required to design with 3/0 Cu with no credit for the lower ambient temperature.
Cheers, Wayne