93K mi 3.0 V6, has a later replacement auto trans. Car performed flawlessly for 24 straight hours at 70-80mph of steady state fwy driving. We took it on track and drove at 1/2 throttle max (in 3rd not D) with lots of accel and decel every lap, but not flooring it-- really babying it for the most part-- and it would just stall at about 40 min and die. restart would idle rough but any application of throttle would kill it. Sometimes we could get it to idle across the parking lot without dying but not always. If we came in and it HADN'T stalled yet, it would restart and be smooth idle, but then die pretty quickly like it suffered from heat soak during the time we performed the driver change. After about 15-25 min, it was fine again and would fire up and perform just fine for another 40ish minutes and then repeat. If it didn't stall while on track it would stall while the next driver was trying to pull out onto the track. While it never gave evidence of overheating (no boiling of water or blowing off of radiator cap to the overflow) the temp gauge didn't work properly (tried 2 different 1-wire senders --needle always read to the first or second line max) and the fan never turned on unless temp sender was unplugged or with our 4th wildly varying 2-wire sender) .We had 2 used 2-wire temp senders and 2 new ones. One of the new ones was about 15% different in ohms than the 2 used ones. The 4th one was wildly different spec and from Rock Auto and part # says it's for a caravan. Anyway, that 4th one would turn the fan on even if the car was only idling a short time, but when the car was cold it wouldn't idle for crap. I would plug in the old sensor even though not threaded into the block and it would idle smoothly when cold, then when the car was a little warm I could plug the new oddball sensor that WAS screwed in and it would run smoothly and the fan would also cut on. We never ran with the sender at all beyond the initial test described above because of the stumbly cold running. The problems reported were with using the 2 used senders and the 1 new sender that was 15% different. Another thing we found is a slight dead spot in one portion of the tps sweep at about 1/4 throttle. The car didn't really react to it, but we could find it with the multimeter. Things we did that we thought was fixing it but in hindsight it was just the time off track that allowed it to get fixed: cleaned the grounds under the battery and up by the alternator. Replaced the fuel pump relay (we discovered it was weak/intermittently working on the signal wire side). We replaced fuel filter, dropped the tank and inspected to see if it was gritty in there (it was clean). Checked the charcoal canister purge valve up on the passenger side-- it seemed to be fine. We suspected the ECU was getting warm (it WAS warm to the touch sometimes) so we removed driver's side headlight and put a protruding scoop for air and even put the windshield washer line into the intake so we could give it a squirt from time to time into the stock CAI to cool the ecu. We didn't have a fuel pressure tester, but when the car was working right we could pinch the feed line and feel the increase in pressure when blipping the throttle. But when the car was not running right, we couldn't blip the throttle because it would just cause the car to WUH -UHhh or want to die / choke out. Fuel tank DID get warm-- well, warmer than the chassis metal under the car. After the event, the car performed flawlessly driving first for 30+ min steady state on the fwy, then about 20min of stop and go, then another 20min steady state on the fwy. Totally flummoxed. Any ideas on what could be going on? Car is not near me currently and it's fate is unclear (currently listed for sale on CL in Houston but I may end up flying out and driving it back to SoCal). I'm just looking for a stack of possible ideas and a means to check or diagnose them for myself or to pass along to the future owner(s). Oh, and there was no CEL and the only time we got one or could pull codes was when we unplugged the temp sensor.