Take for example a 2" diameter cylinder with a 1" diameter rod. The piston blind side area is 3.14 in2 and the rod side area is 2.36 in2. So for our example piston the volume ratio of the fluid in the rod side is 3/4 the volume in the blind side.
For a second, pretend there is no load connected to the piston rod. The larger area of the piston blind side will result in a larger force when both sides are at the same pressure. The piston will move in the direction of the larger applied force. The hydraulic fluid from the rod side must flow out, and the regen block redirects that fluid back into the blind side of the cylinder. This means the the pump doesn't have to deliver all of the fluid necessary to move the piston. For every cubic inch of fluid the pump delivers, the cylinder blind side receives 1.75 cubic inches.
OK, now so much for oversimplification, because a piston that doesn't have a load attached isn't much good for anything.
A regenerative circuit does not guarantee that you will be able to move your load, or that you will be able to operate at 1.75x the speed. Force applied is (pressure*blind side area)-(pressure*rod side area). A regenerative circuit operates at a greatly reduced force. Our 2" piston with a 1" rod can deliver no more than 1/4 the force of a non-regenerative circuit (usually slightly less when we account for pressure loss, friction, etc). The reduced force reduces your ability to accelerate the load (accel=Force/mass). If the load is very light (eg, easily accelerated) and the stroke is very long then then a regenerative circuit may pay off. If the load is high and the stroke is short then its doubtful that a regenerative circuit can actually move faster. You'll have to do the math for your particular system to determine if there is a pay off. And unless the payoff is large, eg, a light load with a long stroke and less than precision control is OK, its probably just not worth the hassle on a servo system. Personally, I'm not a fan of regenerative circuits but they do have their place. Last week I was training Bubba on a hydraulic vibratory feeder (very very low force application and not servo/proportional) that used a regenerative circuit to achieve very fast motion with a small pump. I had to explain it several different ways, but I think he finally got it.
Our distinguished colleague Peter has argued till he is blue in the face on other forums trying to get people to understand the relationship between force and actual stroke time. Some people get it. Some don't. Some are hung up on a simplistic approach of using only fluid flow and never will get it. But just take him at his word - he knows it way better than my simplified response above. At any rate, I hope my explanation helped.