2JZ-GTE VVTI, M150, VE reduced by around 12% in high rpm and maximum boost

My engine has done something that I can't understand until now and I hope someone can help me and advise what to check to solve this problem. My car suddenly started taking around 17 percent off in the closed loop to keep the car lambda as per the target, I always made the VE rich around 2 to 5% so the real difference in VE is less around 10 to 12% . This happens in rpm above 4500 and boost around 1.45 bar.I reduced the VE by about 10% to go back to the range of -2 to -5 % when closed loop is working. I don't understand how can VE change without changing anything. Also now the injectors are working at 58% duty cycle , before it was reaching 68% duty cycle at maximum rpm at 1.45 bar. The boost is right as I have a gauge other than the map sensor connected to the Motec M150, also I have done a leak test and nothing is leaking from the intercooler or the intake manifold , I hear only a minor leak from the wastegate ( which I think is normal ). I checked in the log, the camshaft aim and camshaft position and they are almost identical( ex.4.1 can be 4.0). What can be the reason that the VE was reduced by around 10% , VE in high rpm(from 4500 to 7300) and boost 1.45 was around 102, now it is around 92, which seems too low. Also I changed new spark plugs after this issue (Denso IK27). I'm planning to double check the timing of the engine and compression and also I want to get another wideband sensor to double check that my wideband sensor is reading right. Please advise if this is the right way to go and if there is anything else to check. Thank you

Car setup

2jz-gte VVTI , Motec M150 with john reed package

Stock intake manifold

3 inch intercooler piping

stock intake manifold

precision 6766

ID1050x injectors

Have you checked Actual Compression in each cylinder?

I did a compression test but before this happens so I'll make a new one to make sure nothing dropped but is there any other parameters that I should be looking in to. Thanks

Well, less injection duty cycle means less air is going through the engine with the same level of boost. It is wither air leaking somewhere or a restriction appeared. Also could be timing belt slippage...

Thank you for your advice. I have changed air filter with a new one just to make sure there is no restriction , same results. Also checked that the throttle body is opening 100%. Will check timing and compression as a next step. I have changed my spark plugs from Denso IK24 to Denso IK27 ( I don't think this could be of any connection to this issue ) IK27 is a one step colder spark plug than IK24.

Any change in fuel pressure?

Fuel pressure same, timing ok , Compression done and all seems ok . All cylinders from 11.5 to 12 bar. New air filter , no leak in intercooler pipes or intake manifold, throttle body opens 100%, Turbo compressor blades looks good nothing damaged,VVTI for intake cam is working fine, I reach the boost I aim for, seriously I don't know what happened to get a less 10% in ve table in high rpm and also less injector duty cycle. It is a very strange thing but I really cant find the reason.

Is the variable cam position tracking the expected aim? Have you physically checked the cam timing. I had the CAM trigger wheel on a Mazda intake cam slip once (the trigger is just a press-fit, so I now tack-weld those), could that have happened here?

Yes the variable cam is tracking the aim, very minor difference for example 3.1 could be 3.0, but I always had that minor difference , I have physically checked the timing , I'll attach a couple of photos also for timing. I can send an old log and a new one if you are interested to have a look. I have to mention something that I installed before having this problem , water methanol kit (Aquamist HFS3 v.3.1), but ofcourse when I take my readings I disable the kit , I even removed the control unit to make sure it is not spraying.

Could you possibly have a tiny boost leak around the water injection nozzle? do you have a check valve as close to the nozzle as possible?

I don‘t have somewhere near the nozzles to check with it but I pressurize to 22 psi , I should hear it if it is leaking, also it is very hard to have a leak there as the seats are welded to the pipe and then the nozzles tightened to the seats and I have checked the welding and it was good. i really don‘t have an explanation except one thing , when I tuned the car , humidity was around 20% and ambient temperature was around 15 to 18 degrees celsius, now the humidity is around 80 % and temperature is around 30 celsius. I know the Motec M150 accounts for temperature but it can‘t account for humidity, maybe when temperature went up and also humidity, the oxygen got lesser , it is a crazy explanation but I really cant find anything else. Nothing in the inputs accounts for humidity but 12% in VE is a lot and I don‘t think humidity alone can do this. It is a big question mark for me???