One way to look at is the faster you flow, the less time it takes to get all of the mash liquid through the heat exchanger.

Say you have 10g of wort. Is you flow at 0.5 gpm, it takes 20 minutes for the entire wort to pass through. If you flow at 2 gpm, it takes 5 minutes.

I know it isn't that simple, as the mash liquid is continuously being mixed as you recirculate, but it is a reasonably accurate concept.

The simple yet factual answer is, the heat transfer coefficient for forced convection increases with flow rate. Think of a little cylinder of water traveling through the coil. The amount of energy transferred from the coil wall to that small volume is proportional to the temperature difference between the cylinder of water and the tubing wall.

If you slow the flow way down, the temp of that little volume quickly matches that of the tube wall. So, for most of its trip through the coil it isn't doing much to help. If you speed the flow up, it will have a greater difference in temperature for all of its trip, so it will exchange more energy.

The usual conclusion is that because a slow flow means that the egress of the coil is at the HLT temp, then maximum heat transfer was achieved. I guess it was, if you are only concerned with heating up the wort in the coil. But you have gallons sitting in the mash tun, just waiting to get themselves through the coil as well. By flowing faster, you get all that volume through the coil sooner. And as that volume goes through, you are heating up the entire mash, and eventually the mash liquid coming out of the coil will be near the HLT temp.

To summarize a long point, just because you've achieved great heat transfer to the little amount of fluid in the coil doesn't mean you are doing a good job of getting that heat into the mash vessel quickly.

Sorry if that isn't more clear, this is tough stuff to explain. Here's an online calculator where can play with the numbers. Just put in something close to our values, and try different flow rates. I'm not sure it is accurate for our class of heat exchangers, but the relationship to flow rate will be there.

http://www.engineeringpage.com/cgi-bin/he/h_tube.pl
If I think way back to undergrad, the heat transfer course was one of the hardest that we had to go through.