Yes, for example a 50-25 bank would be from the nose at the 5th diamond away from the target pocket to nose location midway between the 2nd and 3rd diamonds on the banking rail.
I usually have the OB one diamond away from the cushion. As you get into the shallow banks, it has to be moved a bit farther back to avoid the double kiss or hitting the collimator.
The reason I do this is practical. I'm not gathering data for science. I want to know how to play the table on that day. Example:
During warm up I place the OB on the 20-10 line about a diamond off the cushion and bank it with a firm stroke. It banks short, hitting the cushion about 5" away from the effective pocket center. I place a coin on the top of the cushion where it hit and repeat. If I hit the same spot, I have confirmed the bank is 5" short. With a little trig, this is about 5.5 degrees, which I will round to 6 degrees.
Knowing that, I want to test my result. So I put the OB back on the 20-10 line, about 1 diamond out, and put the CB on the 20-10 line, maybe 18" away. To correct my aim, I take the 6 degrees of expected pushback and multiply by 0.8 (because the OB is one diamond off the rail) and see that I need to cut the ball about 5 degrees. So my aim point on the OB will be about 5mm off center. I will then shoot, with a tiny amount of outside english (to offset CIT), and the bank should go.
Another example: 20-10 bank, OB two diamonds off banking rail where there is a 20 degree cut needed to send the OB along the 20-10 line. In this case:
pushback correction = 6 degrees x 0.66 = 4 degrees
aim point on OB = 20mm (cut) + 4mm (pushback correction) = 24mm off OB center
I will shoot this shot aiming 24mm off OB center, with about 5mm of outside backhand english. This should send the OB along the modified bank line without sidespin, and hopefully, it goes in.
***************
I came up with some of this independently, but Jack Koehler had the same basic approach described in his book. Some of what he has is over-complicated (I look only at the total pushback, not how much is related to specific components), and his chart for angle correction multipliers is not very accurate. But, the basic approach is sound.