Ahh... this problem drove me absolutely nuts when I first started my tank. The answer, believe it or not, is found in calculus. Your 2 pumps are working in series and thus generating an "Unstable Equilibrium curve". If you analyze the slope field of such a curve you'll see that it is IMPOSSIBLE to Match these flows even with tweaking (via needle valves), I promise you.
Here's the mathematical explaination...
Think of the flow to and from the sump as constant (parallel to the X axis). Your y-axis is flow of water, and your x-axis is time. What an unstable equilibrium dictates is that as soon as the flows from the tank to the sump and to tank from the sump become unequal the Sump will empty (if the flow to the tank is greater) or Overflow (if the flow to the sump is greater). The Curve moves away from the equilibrium constant expontentially! This will enevitably happen no matter what you do. For example a tiny piece of algea will momentarily slow the rate of one of your flows.
This is best described by a graph, if any one really cares I can draw and post what this graph looks like.
What was suggested above was to install an overflow box, this will fix your problem by setting up a "like" stable equilibrium curve.
If you really care to know the math behind why the overflow box works I will post why, but somehow I have a feeling that no one really cares. Nevertheless, even though I fought with it for 3-4 days before resorting to my calculus textbook I found it to be one of the more enlighting situations about how my water movement works.
Good luck with the new tank!
If you want to see any of the Equilibrium curves tell me, and I will try to draw them (on the computer) and post the image.
Pete