Ray,
To illustrate the point, I have once stuck (on the dyno) a camera tube, like the ones used for medical internal imaging, down the throat of a 34 PICT-3 while the engine was at high load and power. What I saw was not pretty. Large fuel drops falling down from the carb, hitting the T in the manifold, and splattering and bouncing off every which way, depending on the airflow direction at the time. Manifold heating in the Type 1 was apparently used not only for cold start, but also to help with the wall wetting problem. No wonder the AFR traces of center-carb'd Type 1s look so "noisy".
There's a reason OEMs are going to GDI. With correct fluid dynamics they can get it so that the spark-plug area sees a richer mixture that's easy to ignite, while the rest of the chamber gets a leaner mixture with EGR mixed in. Also no mixture loss during overlap of course.
The other reason is that injecting into the cylinder during the compression stroke uses the heat rise from compression to evaporate the fuel droplets. Even though gasoline has a low latent heat value, the latent heat energy taken out by spraying fuel into the chamber lowers the pre-spark temperatures enough to allow higher compression without knock. That effect is much more pronounced with alcohol fuels, as ethanol and especially methanol have very high latent heat values.
There's an MIT paper on that and Ford is building an engine that uses it: Direct alcohol injection combined with port gasoline injection at low loads for a turbo engine. The MIT guys calculated that a 12:1 static compression engine, with peripheral direct alcohol injection (high swirl ratio) can run at up to 4 bar of boost without detonation:
http://lfee.mit.edu/public/LFEE%202006-01%20RP.pdf
That's also the reason I would never do water injection pre-turbo or in a plenum. Getting the water as much as possible in droplet form into the combustion chamber uses their high latent heat value during the compression stroke. This lowers the pre-spark temperatures, and thus the knock threshhold, much more than diluting the charge with water vapour would.