I can't tell for sure from the pictures but your robot may have two major problems that Predator's original Iron Willie had.
The minor one is how stiff the bridge is. Too stiff a bridge too close to the cue ball can affect how much squirt you get. With a long bridge, as shown in the pictures, I think this is probably not major.
The more important issue is the stiffness of the coupling of the "grip hand". If this is not roughly the stiffness of a human grip the results can be affected in major ways. In the case of the experiments in the Jacksonville Project this problem was discovered/understood late during the week of testing and probably polluted some of the results.
The solution was to use bubble wrap between the cue and the machine at the grip or to have humans shoot shots where the effects were objectionably large. I have heard that Predator has since fixed both of these problems. (Iron Willie Sr. also had a very stiff bridge fairly close to the cue ball.)
Hey Bob, I hope we can stay in touch. I would love to have all your input on the testing that we do. It doesn't matter if its wrong or right as the tests will prove the point. But like in a pool game it is nice to look at the shot from all angles. So any theories you have will be greatly appreciated and will definitely be considered in all the testing.
To clarify your statements above the robot proved you are right.
Too stiff a bridge too close to the cue ball does affect how much squirt you get. For the record are robot is adjustable but does allow us to test this point. I would also like to add that even for the robots that are not adjustable that the squirt or cue ball deflection should be the same for all the cues tested in that robot. The bubble wrap is a good idea but since it is not the same thickness to the thousands and because it has to overlap
may through off the results as it would not be the same for all cues. The bubble wrap could move or get squished more with one cue than the other. Our robot has a special machined rubber that is the same to the thousands and is very consistent. The grip is adjustable and well as the wrist joint. How ever my robot does not drop the shoulder. As many teachers say the only moving part should be the elbow. It would not be to hard to make a second robot that does drop the shoulder. The shoulder could have springs that control the dropping of the shoulder. Our robot also has an adjustable bridge meaning the stiffness of the bridge is adjustable as well as the bridge length and the follow through. This was designed this way to be able to test many different theories.