You don't say much on the history of this device and why it is being repaired. Logically speaking, if PP3V0_TRISTAR is giving you 0V and you don't measure a short on that rail, then it is more likely to be a problem with the PMIC (U1202). The same applies to PP1V8_SDRAM.
However keep in mind that the PMIC may be current limiting the output because of shorts on the other rails. Ideally, you want to test the main rails it produces before replacing the PMIC. I would double check my measurements and in the case of PP1V8_SDRAM, I would even consider removing L1216 to "break" the rail. That way I could measure on both sides to see if the problem is upstream (PMIC) or downstream (Tristar or Mesa Boost).