Clipping the wheel is a solution but only just as its still a band-aid as well, and can only do so much. Small housing with turbine wheel design is the issue. It depends on blade count, angle, and weight. for the gas flow quantity, or how much fuel/ air you are pushing with respect to housing and turbine wheel size.
You probably need to see or test what EMP you have first. If its over 2:1 then there isn't anything you can do, short of fitting a larger AR housing for that size turbine wheel for your fueling tune level.
This stuff is a science, trial and error will only go so far. A 18G is a small turbo for a TD42 with a 12mm IP. You have stated you have full spool value by 2000rpm, which suggests you have fueling to reach the volume gas capacity well before 2000rpm so you have reached the capacity of the turbine wheel well before 1800rpm. So you are trying to gate off that excess after 1800 rpm but by 2500 rpm you have reached the limit of what the gate size and actuator travel design will allow to flow. Yes you can play around with springs and get a slightly better result but still you are restricted by the spring and gate flap design.
I will try and explain this concept best i can. Say we have a 30psi spring or a spring with 30psi suggested tension for a actuator. we have set the actuator so it has a tension on the gate flap to crack open at a boost pressure of 30 psi but our EMP at 30 psi boost pressure is 60 psi. So if our gate flat hole is 20mm dia we have a surface area of about .5 inch so we have 45 psi acting on the spring if mechanical advantage on the gate leaver is 1:1 which it isn't the flap arm is about half the length of actuator leaver, so we have a extra mechanical advantage on the spring so 60 psi acting on the spring to hold it closed. But soon as the gate opens that EMP pressure acting on the spring is reduced so the gate will open an amount that is in a equilibrium position with EMP and spring actuator tension. So as more exhaust gas is increased with rpm the higher the EMP will go. The point is here the housing /gate sizing with turbine wheel has a set amount of gas it can flow or PHI so playing with gate settings and tension can only do so much. You have to change the housing or turbine wheel size or design to get control back on the gate.
As a side note boost pressure isn't a measure of a turbo and its flow capacity. This is where design comes into play, it is possible to have a turbo that can flow better volumes in spool with a lot less boost pressure so you get a higher torque rise, air quality or density with fuel is what makes power/torque. By the same token super fast boost rise isn't a measure of driveability or torque. For most of us drivability is the single key to a nice engine. Sometimes or mostly you have to compromise some very low torque to get flexibly in the rpm range. Its usually a engine that can develop its power at a constant rise all the way through its rpm range is the nicest to drive. Torque is a big factor on how you get that power curve, as power is a product of torque, not the other way around. You measure torque to calculate power, power is a math calculated value.
EMP above say a suggested 2:1 EMP:IMP will kill driveability it reduces power development after peak torque. Too small of a turbo will show up on a dyno with that characteristic power flattening off and even dropping off after 3000 rpm.
I find for a good powered sub 200rwkw TD42 you need a comp wheel size of about 67mm trim dependent and a turbine wheel size of about 58mm trim dependent to get enough flow to get gate control and more respectable EMP:IMP ratio. But this EMP stuff isn't the be all to end all either its just a big consideration in the big picture, as is everything else in the turbo sizing game.