I wanted to try that as well back in the day but never got around to it, looked like it would work even better than the 2.0 manifold deal, which worked pretty good.
And, just to make a point, you can have vastly different lengths in intake runners and them still work pretty good if basic shapes are right. Just because two are short and two are long means nothing sometimes. Vizard checked the 2.0 intake and came away with around 93% efficient all passages considered as far as the flow bench went. Sure you'd like them all to be the same but the real world comes knocking. Edelbrock made fortunes by using different size passages on their X style manifolds to make up for differences in length. It's not the length, rather the VOLUME of the runner that counts. Look at a Torker manifold, none of the runners are same length and all will be different size passages as well, the shorter ones will be bigger and the longer ones smaller and port them to all the same internal size and watch the manifold lose 25 HP faster than spit.
I have no idea what one means by a 'waterfall effect', the D port manifold works by keeping flow at the port roof where most of the action is at that point anyway, putting it on a D port head should be even better. The lower 1/3 of a 2.3 intake port entry is pretty much dead air anyway and the head filled there will speed up action in the rest of the port. And why D ports often work anyway. When the relatively dead slower air in bottom of port gets to the restriction at the valve guide it tends to interfere with the upper faster flow to make for turbulence. Cut it off so all is rushing at closer to the same speed and then it all gets around the turn faster and smoother.
Pretty much a waste of time flowing a header, you are testing there in a fashion that does not imitate the real world, exhaust first cuts loose in the pipe like a stick of dynamite, with pressure rushing through at sonic speeds. No flowbench on earth can test at close to those conditions. The choice of pipe sizes and close to equal length will have far more impact there. A flowbench cannot reproduce the exhaust gas plug effect firing down the pipe at all.