*EDIT: Now I am happy with up to 24psi, but really depends on fuel flow and if intercooled or not and should be tested. I am happy to talk with each individual and assist with self tuning.*
Boost and air temps.
If a gas in compressed, the temperature will increase if there is not enough capacity to remove it somehow. The old formula (P1V1)/T1 = (P2V2)/T2 (T is always in Kelvin by the way +273.16°C)
However the compressor is not 100% efficient either, and the energy lost due to inefficiency serves to increase the compressed gas temps higher than theoretical. An intercooelr serves to mitigate both issues and makes a substantial difference in most applictaions.
The modified turbo, which uses several different components and is completely new internally is absolutely fine at 21psi.
The 4AGZE being petrol has nasty high EGT's. I just bought an MR2 turbo off eBay to supply me some parts and the turbine shaft was dead due to excessive heat. This is the second MR2 turbo like this that I have seen. These smaller capacity engines work alot harder and the EGT's are typically even higher than for example the 7MGTE (3 litre). I bought 2 very low km JDM 7MGTE turbos ages ago in near perfect condition. However, the shafts are no where near as good in condition after ~ 100,000km on the 7MGTE compared to 12HT turbos I have rebuilt at 530,000km! The 12HT shafts are almost always perfect. After a clean up they totally look new. The original 3SGTE MR2 Gen 2 turbo I dissasembled on friday measured 9.67mm on the turbine side bearing where it should be 10mm! That is unrepairable since you cant buy undersize bearings for the CT26 (they have to be custom made).
In any case, long story short, I had absolutely no problems running my high flowed CT26 on a 1HDT up to 27psi. It had a 270 degree thrust bearing. Nowdays I change that to 360 dgree mostly, especially for manuals but after 50,000km and 4 years, it runs totally fine still. I no longer own the car but a relative bought it so I keep contact.
Everyone is free of course to excercise their own levels of caution and I reccomend to do what you feel is comfortable. If I didnt push the boundaries of things and be ready to suck it up if it fails, I wouldnt have something to offer you. I have only seen one example (not personally, just photos) of a blown apart CT26. If it was caused by overspeed and not by intial contact of compressor wheel coming into contact with the compressor housing, I can only suggest that it was run at choke flow and had dramatic overspeed. A high flow turbo is far less likely to get to that. The way to blow up a CT26 is to disconnect the wastegate on a CT26 with a 12HT/1HDT/1HDFT stock compressor wheel and turn up the fuel delivery. It will smoke up top, egts will be high and after 2500-2800rpm it is on the choke line. Then rev to 3500 and it will probably blowup.
I ran my wastegate disconnected for 30,000km's BUT it had a high flow compressor wheel so didnt hit choke flow.
These are a well designed turbo and very strong but very conservative trim on compressor wheel and turbine - an older philosphy for tuning. I would suggest they are overbuilt given a stock CT26 has a 10mm shaft whearas the coveted TD06 mitubishi that can do >550hp on a petrol has a 9mm shaft. also the CT26 shaft is longer between bearings which gives geater rotating mass support.
21psi is the most I would reccomend on this new turbo (depends on engine you have and rpm range and intercooling etc). This limit is not because I consider damage will happen (it might, but I dont know), it is because it might hit choke flow at higher rpms. It has been run to 1.8Bar in testing ~ 2000rpm but I cant possibly reccomend it.
It is not the answer to everyones requirements, but if you want low rpm performance and more boost up top than stock, it is one answer and the only bolt on solution I have come across in 4 years looking for one!