I worked for a company, now gone, that was developing a launch vehicle for geosynchronous orbit years ago and the engines were pressure fed. We developed tanks which were polyethylene lines carbon filament wound with aluminum polar bosses for connections, single use. some of these tanks were for RP1, some for 92% H2O2 and some were for high pressure helium (a noble inert gas which would not react with the H2O2) for use in pressurizing the system to feed the oxidizer and fuel into the manifold for the engine. the helium tanks were intended to pressurize to 4500 PSI and we hydro-tested to 1.5X that. We had this test fixture to hold the test article (the tank which was a 1.5 meter spheroid) which was made of 6 legs of extruded steel tee sections roughly 3/4" flange and web 8"wide x 6" deep with a top and bottom of like construction. The test crew and engineers took the article to full hydro-test pressure then drained it and began to pressure it up to 4500 PSI helium; watching the video in slow motion of the event was interesting, they were coming up to 3000PSI when a sound, best described as 4 metal guitar strings pulling into, followed by a catastrophic explosion destroyed the test article along with the test fixture (turned the test fixture into a small patch of large steel gravel).
In my humble opinion, the data may vary but the results are the same, catastrophic failure; and I am fairly certain for much the same reason; layers of carbon filament laminated together with resin to form a vessel of a relatively large diameter do not take the cycling of low to high to low to high to low pressure and de-lamination occurs. With all of this said, last I heard they are making some 2500 to 3000 PSI pressure vessels out of Carbon fiber now which are holding up but to my knowledge none of them are over 20" in diameter which helps with the cycling and the pressure issues.