Oh I know, it's not gonna happen and that was my point. And I really don't want to start the tubular vs. weld-to-floor subframe connector pissing match again either, it's been beaten to death over and over again. Truly, until someone builds at least 3 or 4 A-bodies that are identical in every way except for the type of subframe connectors and donates them for destructive testing and analysis the question won't get even get a good preliminary answer. Because really you'd need dozens of cars and multiple tests to come up with a definitive answer (just like the safety testing done on modern cars now). Even a full 3D solid modeling analysis would only be a start, you could always argue about how all the different welds were modeled, what assumptions were made etc.
And since none of that will ever happen, we'll continue to see people make assertions on which type is better without having any facts or evidence on which to base their assumptions. There are too many variables in play to guess at which is "best" or "strongest". And given the complexity of the problem it would probably come down to application anyway, because which type is "better" would likely go back and forth between different tests. One kind might do better for torsional rigidity, the other might do better in pure bending, all of it would depend on where the loads were placed etc. Heck it might even be different by bodystyle, 111" wb vs 108". And at some point you would find that it doesn't matter how stiff the subframe connectors are because the bending would just happen somewhere else instead, there's a limit. At some point the chassis would stop improving even as the strength of the subframe connector was increased, because a subframe connector can only do so much.
So, install whatever type of subframe connector you can. The contoured, weld-to-floor kind are much easier to install on a car that's already been fully stripped, blasted etc, and put on a rotisserie. If the entire car isn't being blown apart and taken to bare metal the tubular kind are much easier, and even someone with fairly limited fabrication skills can make them at home if costs are an issue. Both kinds make a noticeable difference in the kind of anecdotal things we can "test" ourselves, things like how much the body gaps change when jacking the car, the sound the doors make when they're closed, how many squeaks and rattles the car makes on the road, and on the higher horsepower cars whether or not cracks start to develop (or re-develop) in the paint at the quarter seams and other problem areas where flexing occurs. Those are all legitimate ways to tell something changed, but none of them are a good way to compare one kind of subframe connector to another, the difference is too small between the different kinds and the results are too subjective.
