Ernie, sump size becoming a problem is dependent on your overflow to sump capacity. If, without the surge device, your tank to sump to tank to sump loop is being run based on the limitations of the overflow pipe diameter/flow capacity (terminal velocity/flow), then the filling of the sump is limited by the rate of the overflow. Therefore, in general, the surge device will not significantly fill the sump anymore than it already is, it will however, fill the overflow box. Water level in the sump/overflow to the sump will be increased slightly or significantly by the increased head pressure of the water standing on the return pipe if the overflow pipes are not being run to their maximum flow capacity. The increase in flow will be dependant on how close to the terminal flow-to-sump capacity you already run the overflows prior to the surge device installation.
Rate of surge will be dictated by two main principles: Gravity and constriction. You can have the surge device as high in the air as possible. It will flow/siphon faster the higher you go, to a point: the maximum flow capacity for the diameter/length of pipe you use as the siphon tube. Ideally, the largest pipe reasonable for your reservoir should be used. This is limited severly by practicality. Using a three inch pipe as the siphon requires a large amount of head pressure/volume over the top of the siphon bend to prime the siphon. The more overhead pressure you have, the more volume the surge is going to be, again, causing overflow and sump limitations to figure in.
I'm not familiar with the overflow volume in the tank you have, however, for practicality's sake, the surge device volume should be limited by the volume of the freeboard left in the overflows (ie, volume of overflow minus the water volume normally present) + the volume of any remaining freeboard in the tank itself (unless you're running monster overflows and a big sump) + the volume that will flow to the sump during the duration of the surge device's delivery of the surge (compensated for the varying rates: ie, for every second, the surge device delivers x amount of water, and the overflows deliver y amount of water to the sump (this last figure to add is only effective if your overflows are not being run at full capacity). This one can be hard to figure and you start to get into conjecture, so be conservative. If you feel so intrepid, you can do frictionless calculations of gravitational acceleration of the mass of water; mass of water will be determined by volume of pipe (diameter and length). This should give you a nice conservative number, because, even so, for an accurate figure you have to add to this time from backpressure from the submerged outflow, friction from the pipe (unless you can find the friction coefficient of the pipe), etc.
On another note, don't short change yourself on the surge periodicity, make sure the overflow can totally recover. Also, don't forget, whatever you are returning to the sump via the surge, the sump was already taking a deficit from whatever was pumped to the device, so make sure the sump has the volume in it to fill the device without running the system pumps dry or cavitating.