My investigation of code 4B99 revealed that the DDE is detecting a short circuit in an injector on bank 2. Not that different from the ISTA fault description. The ISTA "action in service" puzzles me and appears to indicate that the injectors couldn't be the cause of the short circuit.

The guys on this forum highlighted the partition water leakage issue for me and sure enough, I found lots of corrosion around injector #5 and on the feed line.

This all started because the engine would suddenly die when I went over ~70% throttle. Eventually the engine died in the middle of the road with no throttle and it would not restart. Knowing that the #5 injector was somewhat shorted (possibly intermittent), I disconnected injector #5 from Bank 2 and the engine fired up right away. Ran a little rough but I was able to drive 2kms home and avoid a tow.

I then disconnected all the Bank 2 injectors and measured the resistance between the two pins, pin 1 and ground, pin 2 and ground. The resistance between the two pins was consistently across all three injectors (~210Kohms) but only injector #5 measured about 5Kohms between one of the pins and ground. That measurement was ~infinity (beyond my meter's measurement capability) on the #4 and #6 injectors.

And that's why I replaced injector #5.