I've developed quite a bit of FP4+ in both 4x5 and 35mm formats using the same times for each in D-76 1+1. Like you, I'm using an HP CombiPlan tank for the 4x5 and a "no name" plastic tank, similar to the old Paterson model, for roll film development.
My first thought was that you were not using enough stock solution for the 4x5. When I re-read your post, I realized that with D-76 at 1+1, you are comfortably well past the required minimums. So we can rule out early developer exhaustion as a cause. I'm assuming that you have adequate temperature contol for both systems. There is no reason to believe otherwise since you can partially submerge either tank in a tempering bath. That leaves us with only one other possibility, namely insufficient agitation.
My agitation scheme is a bit different from yours. I use the twirling stick with my plastic tank and go through 6 complete back and forth twists in 5 seconds every 30 seconds. Granted, that's a fairly vigorous agitation scheme and a lot of people are not comfortable with that. I don't agree. Vigorous agitation never hurt anything. The rest time between cycles is the variable that controls the rate at which development proceeds. It doesn't seem to matter if the rollfilm tank is overly full with this technique either; development proceeds evenly. Agitation with the CombiPlan tank follows the same timing cycle, though with five or six alternating rapid inversions in 5 seconds each thirty seconds. The tank is flipped once to the front and back, then to the side and back. Each flip and back is counted as one inversion. It is important to not fill the CombiPlan tank to the brim. You need some head space above the developer to insure a complete exchange of old developer for fresh when using an inversion agitation technique in any type of tank.
My suggestion to you is to give your CombiPlan tank more vigorous agitation and stay with the same times you've been using. See if that doesn't clear up the problem. If not, simply extend your development time in small increments until you get to your desired contrast. Change only one variable at a time or else you won't be able to account for any changes, desireable or otherwise.