The thing about articles and charts like this-- they're super-interesting, and can lead you in a lot of directions that may (or may not) work in the real world.
@RS-60 mark brings up probably the very first thing a guy will find if he reads a lot of engine esoterica-- that according to the formulas, most of our engines are over-carbureted in mathematical terms. The thing is, guys who build Type 1 VW engines for a living know what works and what doesn't... and almost all of them will say something to the effect of, "yeah, the math is nice- but these engines run better and make more power with bigger carbs than you find in the charts". Obviously, there are limits, but a 2276 with 45 DLRAs is not over-carbureted.
I had a friend once who explained the smaller carb thinking like this: if the venturi in the carb is not the choke-point for the intake, then your air-speed is not going to be great enough through the carb for fuel to be properly metered. In other words, unless your carburetor venturi is smaller than every other point along the intake (intake manifold, ports in the head, back of the valve, etc.), we're all fooling ourselves with bigger carbs. It makes perfect sense when you think about it, and has almost zero application in the real world.
I have no idea why it isn't so, but I can tell you this without any hesitation- an engine with very big ports and valves in the heads, and undersized carbs is a dog. When I tried to de-fang my 2332, I ran 40 DLRAs on it for a while with great big heads. Theory aside, it was a pooch. Airspeed plays a huge role in making power on the street, and the holy grail is getting flow numbers with smaller ports (which is why the Panchito is a near miracle). It takes a defined amount of fuel to support a given HP (assuming the fuel stays atomized), and super-small carbs can't provide enough to make the kind of power I've grown to expect.
I talked with @Pat Downs this week-- he's been getting 180-ish HP with 2332s running Panchitos (with 40 mm valves, and size-tiny ports) and bigger (CB's 2292 grind, which is a lot like an FK10) cams with a lot of overlap. My engine is out of my car, and I'm planning to step away from running a science-fair project twin-plug set-up for my car (more later as to "why"). Running a bigger cam with smaller heads and lower compression flies in the face of every formula or chart I've ever studied late into the night. But dyno numbers cannot be disputed, and the power he's making with this combo is astounding.
I'm going to give it a try- running the heads I've been high on since they came out and a cam at least one size bigger than I ever thought was reasonable. This time I'm building what a seasoned professional knows works, rather than what my theoretical mathematical chart tells me might be better.
Time will tell, but I'm trending in exactly the opposite direction from where I've been, and I've been lost in the math for a long, long time.