This is a wonderful example of how theory doesn't always work so well in the real world. Two "identical" cylinders given "identical" inputs and loads should respond equally, right? Well, not so much. There are many sources of "inequalities". The layout and condition of supply and return lines and connectors. Minor manufacturing differences in the cylinders themselves. One of the biggest ones though is the absolute impossibility of having the operating axes of the two cylinders in perfectly parallel alignment. That condition causes side loading, which in turn creates high friction loads, which in turn affects linear positioning. It is always a bad idea to rigidly connect two or more cylinders. You can guide your load, but the connections to the cylinders should allow for some flexibility and misalignment.
There is one theoretical way to connect any number of cylinders and have their motion absolutely coordinated, but it has significant "real world" restrictions. First, you have to use double rod cylinders so that the pressure area is the same on both sides of the piston. Then you plumb them in series: pump to inlet of cyl #1, exhaust of cyl #1 to inlet of cyl #2, exhaust of cyl #2 to inlet of cyl #3, and so on.
The best way I have found to synchronize the motion of separate hydraulic cylinders is with a hydraulic rotary flow divider. (Google it.) This device uses the same impellers as a gear pump. Two (or more) impellers are mounted on a common shaft. They operate inside a close-fitting housing that basically creates small enclosed volumes of hydraulic fluid between each tooth of the gears. Two gears operating on the same shaft have two identical flow rates.
My experience with these devices is pretty good, but they do have their limitations (long stroke inequalities, etc.)