Do you have evidence to back up that claim? It seems counterintuitive that slowing down the top few percent fastest drivers would improve traffic. If there's enough traffic to cause congestion wouldn't they be stuck behind slower drivers anyway?
As you go faster, your stopping distance (that is, the distance you should maintain between you and cars in front and behind you) increases. This leads to roads having a maximum capacity somewhere in the area that speed limits are set. If you go faster than that, you're reducing efficiency of traffic.
Anecdotal evidence, but it also makes intuitive sense if you consider that vehicles travelling at a range of speeds and particularly changing lanes to overtake the slower drivers is effectively a turbulent flow. If that flow can be smoothed, as a body it travels faster.
I know the trucking industry regularly criticizes the safety implications of having a separate speed limit for them, as it makes lane changes much more complicated. They aren't impartial to the financial impact of the speed limit though, so there's that.
