It depends upon what you are trying to accomplish. If your reasoning is that by restricting the volume of the barb, you will increase the velocity of the water passing through it and hence make it strike the baseplate with greater force, you may be disappointed with the result. For the same driving pressure, by restricting the volume, you will indeed increase the velocity at the sacrifice of the rate at which the water moves through the block. The trade-off can yield a positive reduction in temperature of the area where the water strikes the baseplate, typically, directly over the part where it rests upon the die, however for the heat that propagates laterally out from that area, into the rest of the baseplate, the lower rate of flow may negatively impact the ability to remove as much of that secondary heat from the rest of the baseplate.
To put the question of benefit versus detriment into perspective, lets say hypothetically that 75% of the heat from the die is removed by virtue of the water striking the baseplate in the area where it contacts the die and the other 25% of the heat propagates outward from that area into the rest of the baseplate. That remaining area of the baseplate does not benefit from that initial downblast, but rather gives up its heat as a function of the rate at which the water flows over it. Now lets further say that if you increase the velocity of the downblast and the area above the die now gives up 80% of the heat, but by reducing the flow rate through the rest of the block, it only gives up 20% now, you will have broken even. If it gives up less, you will have negatively impacted your blocks performance.
The point is, without a lot of experimentation, it is hard to speculate where the "sweet spot" is for a given block design in considering the importance of striking velocity versus total rate of flow. That is what makes the whole experimentation process so darn interesting. For me, the redundant nature of "cut and try" is not as burdensome as it is fascinating, so I don't mind taking that approach. Having a patient disposition helps alot. Thank goodness I do not do this for a living because I'd probably be forced to conserve effort by mathematical modeling and to me that would remove some of the "magic" and hence personal interest in the process. For others, the opposite may be true.
Obviously, in the typical water cooled PC, there are many other mitigating factors. Cooling water temperature is a big factor but equally important is your choice of pumps. Not based solely upon some often subjective claimed rate of flow, but by the amount of "Head" the pump tolerates. The pumps we most often choose can only tolerate a certain amount of resistance to flow before the water starts slipping back past the impeller, much like air slips out the back of an axial fan when you place resistance to free air flow in front of it. At the beginning of this now lengthy response, I said "for the same driving pressure". That pressure is not the same for different amounts of flow resistance and adds yet another variable to the experimentation process.
So, go ahead and try different striking velocity techniques, but you better have good die temperature measuring facilities to be able to meter whether you are moving in a positive or negative direction with each physical change. Unless you're real lucky, small incremantal changes will not always yield a large enough performance change to register with less accurate die temperature metering techniques. Above all, keep good notes.
Hoot
P.S. Does anyone besides me think that it's about time to move this thread to the water cooling forum?
I'll probably submit an article on the process for the front page when I have some disposable time to write it up.
