Are your resistors fully operational? When I first got my 1009 I installed all new GP's, replaced the relay and did the card bypass with a momentary on switch, I poked around with a multimeter and thought everything was good to go yet she just wouldn't start unless it was really warm and sunny out, and even then it required a lot of cranking. I installed a new fuel filter and bled the system, and contemplated replacing the injector pump since it was dribbling fuel. After banging my head against the wall and rechecking the GP system for the umpteenth time I realized my resistors were not resisting and I was just blowing out my glowplugs. I decided to bypass the resistors (there's a thread on that somewhere on here) rather than try and find new ones. Once I did that, she fired right up! Why I didn't catch it the first 20 times when following the TM troubleshooting guide I'm not sure, I swear I ran the same tests a million times, but oh well.
As a side note, my typical starting routine involves depressing the pedal while the key is on and glowplugs are warming, but not holding down the throttle during cranking. This method seems to be reliable for me in starting the old beast, even when she was buried in snow this winter, and it seems to effectively set the idle once warmed up.
Sidenote 2: During this whole startup issue I did resort to using starting fluid twice. My method was to spray the diesel starting fluid on a rag and have someone hold the rag in front of the air filter opening while I cranked the engine. Both times it worked, and caused no problems. Both times I was stuck out in the middle of nowhere with no other real options for how to get home so I deemed it was worth a shot. I disconnected the GP relay both times before attempting, not knowing that they weren't doing anything anyways.