You're still missing the point.
You set the shock overall length with the wheel stuffed at full bump.
If you adjust the shock height any longer then you are giving up potential uptravel.
If you adjust it any shorter you risk tire to fender(or whatever) contact.
You set the shock length with the lower ring and leave it.
You run the upper ring up or down wherever you need to set proper ride height. All that setting the shock length first does is ensure you have as much usable bump travel as possible.
You will only have as much useful droop as your spring at full extension minus whatever it is at ride height.
Also like I said, if you're corner weighting using the lower collar, your bump stops will be hitting at different times left vs right.
Having "loose" springs does not matter. You don't want preload unless your rates are so soft vs your available shock travel and lengths which may dictate it. If you're running higher spring rates and they're preloaded you are giving up some of the little bit of droop travel you already lack.
Having large amounts of preload also does other weird things because if you have a 10k spring preloaded 5mm, that means it takes an initial force of 50 kg(110 lbs) before the spring even starts compressing (in the case of a tire leaving the ground or having very little weight on it. This means the tire contact patch gets "shocked" (pardon the pun) with an undamped 110 lbs coming down onto it. Of course again this is all talking in 1:1 MR for simplicity sake, but applies to the front almost directly. The rear requires some math.
Now if you're running a 250 lb spring on a 1000 (sprung) lb corner, it will eat up 4 in of overall travel just to sit at ride height. This is a major problem if you have only 6in of travel like a lot of the cheap Chinese shocks. This means you have 4 in of droop travel, 2 in of bump travel, which will ride like crap and bottom out everywhere (remember this is also without bump stops factored in).
If you preload the spring by 0.5in, it now requires an initial force of 125 lb before the shaft starts moving. So with 875 lb left to support, the shock only compresses 3.5 in, leaving you with 2.5 in of bump travel.
Notice how the exact amount you preloaded directly gets turned into more bump travel (and less droop)? The spring still compresses the same amount, but you're using a different portion of shock travel to achieve it. If you preload it less, it'll lower your ride height, if you preload it more, it'll raise it. (Note that with higher rates the springs will be "loose" at full droop. This is not a problem, though you have the option to run helper or assist springs which is a whole other can of worms to get into).
With rates that low vs the corner weight, I would not corner weight it my method. It needs more shock travel, plain and simple. It's unusable without preloading the hell out of it, or having basically no bump travel. If you preload it by 2in, it'll do all kinds of goofy stuff too - it'll require 500 lbs to even start compressing. Lift a tire and come back down it'll be very abrupt.
I never set my car as low as it could have gone. I figured out my max bump, and based on spring rates and corner rates figured I needed 2 in min bump travel, so set my ride height about 2.75 above max bump based on roll centers (which had been corrected with knuckles, otherwise it'd be another inch+ higher).
This ensured I was running at the proper height, with as much bump travel as physically possible, and because the rates were so high, my ~1 in of sprung down travel was fine, with a bit of leftover shock travel where the spring could unseat. There's no getting around that with a 700lb spring on a 730 lb corner. The tire constraints vs shock length did not permit adding a helper/assist spring to play around with, otherwise I would have tested that.