throttle plates fully closed so no need for choke
Not true. The primary plates are connected via the accelerator cable. Yes, at idle they are fully 'closed', and only move if you press the pedal. However, there is a small 1/16" (or smaller) gap all the way around the plates. This airflow gap is all that is required to start and allow the engine to idle. Even that tiny gap must be fully closed at start up, this is why there's a choke plate. To completely and fully close at the moment of start up.
Do the throttle plates have to be cracked open to start?
See above. They should remain fully closed, resting against the throttle plate stops. And when rebuilding, your books should also say to make sure they're perfectly centered on the throttle shaft (and the respective carburetor walls). If you think of the throttle plates being open during start up in conjunction with the step cams, its just fast idling to warm it up quicker. Same thing can be achieved by barely pressing the accelerator pedal and holding it there for a minute.
Even if no fast idle is wanted? Then, the choke plate is necessary along with the fast idle cam, the choke heat source, and choke pulloff. I assumed the throttle can be fully closed at startup since the plates are fully closed at idle and the idle circuit handles the air requirements?
Your idle circuits is the fuel side of the equation, the tiny bits of air that enter are more related to fuel bleed off or vacuum issues with liquid in a tube (the circuit). (E.g. dunking a straw in a glass of water, and holding your thumb over the top of the straw) The only air side of the equation is what passes through the throttle plates.
Fast idle, the step cams, idle circuits, choke pulls-offs, etc. This is all 'fine tuning' idling characteristic improvements. Think; a carburetor works, but these specific things makes it work better. So lets ignore them all for the time being and focus on the hard starts.
Maybe there is more to the cracked throttle plates + choke + pull-off startup action you described than I gave credit.
The way I understand it, on a warm day, with the throttle plates fully closed, it should start right up after one accelerator pump shot. Maybe it wants some pedal for a few seconds to idle up to over come cold engine friction and some poor fuel atomization but that should be it.
I hate to use the term 'warm day', as even a 90 degree day, a stone cold engine's choke plate should be fully closed. So lets think of an already warmed up engine. It should start perfectly with one pump if you've let it sit for a minute.
[/LEFT]
So here's a test, ignoring everything other than the hard starts when cold. Cover the air horn with your hand as if it was the choke plate. Have someone crank the engine. Does it immediately start? (For the most part). The second it starts, remove your hand. Does it kill the engine, or if it does still run, does it run poorly for a few moments.
Test 2: Wait an hour after test one. Cover the airhorn with your hand again like test 1, have a friend start the engine. Immediately after start up, slightly tilt part of your hand and allow a very small amount of air enter. How does it run? Can you make it run better or worse by moving your hand.
Per the other problem with occasional hard starts when warm, and not wanting to goof with THAT issue by installing the choke plate. If it were me, I'd start with each problem individually, and fix it. If you can 100% fix the cold starts, then move to warm starts, whether it changes or messes with the warm starts or not. That can be addressed afterwards.
EDIT: And after re-reading this, during your rebuild, how snug was the primary throttle shaft? That's a pretty common issue with it being worn out, allowing extra air to enter when its not supposed to.
As far as thinking no need for a choke plate when the throttle plate is fully closed. I know all of us say this from time to time: If GM (or Rochester in this case) put it there, they did so for an absolute and definitive reason. These guys are leaps and bounds smarter than us. 99.9% of the time we have issues with these old rigs, is when we the silly owners try and 'change' things. Every square inch of that carburetor, every part, every clearance, size weight and shape- was designed to perfection. A quadrajet is the pinnacle of carburetor design. Its the closest thing to fuel injection there is. Rochester continued to design carburetors up until the late 80s, when all other designs (Edlebrock, Holly, Daemon, Webber, Stromberg) are all based off carburetor engineering from the 50s-60s. That's why Quadrajets run so well, and that's why they're so much more complex and hard to tune than all others.
For example, in one of the two books, Roe's or Ruggles. There's an entire page dedicated to the various metal linkages. Stressing the point that even the most minute of tweaks of some of these linkages can give you idling problems. That's how perfectly they're designed.
It needs to be there, almost fully closed when cold, you'll have to check your spec sheet, but somewhere around 1/8" open. Its been a while, but something a long the lines of when the engine cranks it snaps fully closed, then after start up it opens that 1/8" from one of the vacuum break linkages pushing it that 1/8" open, and runs that way until it begins to warm up.
