I find it hard to believe that the Y-pipe, coupled to poor flowing manifolds, bolted to poor flowing heads, with a small cam, is THE problem. On long tube headers, with a single wire O2 sensor FAR back from where stock was, I watched my setup go closed loop as soon as the throttle was touched, so there is no heating issue. EGR could be true, but GM also used two different methods of EGR operation, positive and negative, one did require exhaust pressure to operate correctly, so that's a variable, not a constant.
If you look at the variety of charts out there for X displacement + Y HP=Z pipe diameter, you'll find that one half of a TBI engine generally needs 2-2.25" pipe diameter. I'm guessing all the crossovers are right in that range.
From what I've seen on the one y-pipe I cut apart (perhaps a diesel only deal, I don't know) it was dual wall. The restriction wasn't any worse than any other old technology exhaust that was used at that time.
The old cats (pancake style) are definitely restrictive pieces of junk, but that was the technology back then, can't blame GM for that, but it's an easy fix. Plenty of aftermarket high flow converters that don't cost an arm and a leg (if you aren't in CA). I would be willing to bet that on a stock TBI rig, the converter, unless plugged, is no more a restriction than the heads or manifolds. Thus, probably well matched.
Long story short, with long tube headers being worth ~20-25HP on a stock SBC, my opinion, derived from the testing that has been done, is that the manifolds are the major problem. With TBI, head flow will be another issue, but until the manifolds are dealt with, you will deal with that bottleneck. Then you will get into cam, intake, throttle body, etc. You can put effort into a TBI engine and make some additional power, or you can do an LS and get a lot more of everything. Or just leave it alone and realize it's a low RPM torque motor, that's it.