During tip-in (when you initially put your foot down) conditions and during engine rpm acceleration conditions fuel trims may swing anyway... It is arguably more important to get fuel trims correct during constant engine load conditions (steady rpm) than during tip-in conditions. 2nd Gear quickly accelerating to 3500/4000rpm doesn't imply constant engine load. If you're in 2nd gear trying to maintain constant engine load it is likely that the only loads you'll be able to hold steady are low engine load conditions unless you're going up a very steep hill (or riding the brakes lol).
I would start by entering some base rpm correction settings that are equal on both banks ranging from around -3% at 500(five hundred)rpm to around +8% at 6000rpm. Better to calibrate at around 2000rpm than at 3500/4000 rpm because the multiplier makes more difference than RPM correction and at 2000RPM you can still see how fuel trims compare between running on petrol and running on LPG at most pinj points whilst holding a constant engine load. Make sure it stays running in closed loop mode when doing any fuel trims comparisons... fuel trim figures are meaningless if it is in open loop mode, to set up correct mixture for open loop compare lambda readings.
You will always be short of being able to address one aspect I'd ideally want to be able to compensate for on the Jag engine - the difference manifold pressure makes to effective petrol pressure (if petrol pressure remains the same in absolute terms then if manifold pressure rises effective petrol pressure is reduced. The petrol ECU knows to compensate for the effect of fuel pressure versus manifold pressure but the LPG ECU doesn't know that a given X millisecond long pulse of pinj doesn't always supply the engine with the same amount of fuel), because you can't dial in any vacuum compensation using the Stag ECU. But that is only a minor aspect and you should still be able to get good results. More in depth... for any given pulse length the amount of fuel petrol injectors would supply depends on fuel pressure versus manifold pressure. If we assume petrol pressure stays the same and is 4 bar absolute then if the manifold is at absolute vacuum (it never would be but it could reach as low as around 0.2bar) the pressure would be 4 (3.
bar. But at full engine load when the manifold is at atmospheric pressure (1 bar) the 4 bar of fuel pressure becomes 3 bar relative to manifold pressure. To compare the difference in flow we do square root of 3.8 = 1.95, square root of 3 = 1.73, then we find the percentage difference between 1.73 and 1.95 which is around 16% (going up from 3 to 3.8 bar) or is around 18% going down from 3.8bar to 3 bar. On first consideration it may seem that there is some self compensation going on for difference in fuel pressure and to some extent there is... but without compensating for vacuum the tip-in mixture is likely to be further away from ideal than necessary and correction for RPM won't be as intuitive because for any given manifold pressure the engine will breathe a different amount of air depending on rpm (and breathe most air per induction stroke and therefore need most fuel to go with that air) at the engine's rpm torque peak.