It might facilitate better replies if you could provide the pertinent measurements in a format easy to decipher in relation to your other observational comments. If I understand your original statement correctly, seating bullets to book max COAL puts them well into the lands. That part was very clear. But you also stated that "seated with no touch they are above SAAMI minimum COAL". To me, that means you still have room to seat the bullet where it will be jumping (i.e. off the lands) without seating it ridiculously far down in the neck. "Above SAAMI minimum COAL" means within the suggested range, or am I not correctly understanding the situation?
If this is the case, it suggests to me that the simplest explanation is that the freebore is simply too short for the bullet/load you're trying to develop. There are several possible ways around this:
First, I would carry out a seating depth test starting with the bullet seated at approximately .003" off the lands (i.e. jumped, but just clear of "touching"), and moving it out in .003" increments to maybe .020" or .030" off the lands, or at least as far as you can reasonably go before it becomes seated "too deep" in the neck (i.e. before the loaded round goes
below SAAMI minimum COAL). If you find a good seating depth window within that range, it seems like the problem could be solved fairly quickly.
Second, if one ever pulls any commercial .223 Rem loads, it is obvious that the bullet is seated well down into the case with many of them, even those with light bullets. Although we would never choose to seat handloads this way if we didn't have to, the commercial ammunition can shoot just fine with the bullets seated relatively deep in the neck. If one is ever forced to work up a handload under similar constraints, selection of a very fine grained powder that has a slightly faster burn rate than might normally be selected for the chosen bullet weight can sometimes help. Nonetheless, it is possible to work up a solid load with the bullet seated pretty far down in the neck (i.e.
below SAAMI minimum COAL).
Third, would it not be possible to simply optimize seating depth with the bullet seated into the lands? I generally do not choose to seat bullets into the lands unless I have to for several reasons. Fortunately, the bullets I most often work with do not insist upon being seated into the lands. Regardless, shooters in some disciplines rarely ever shoot bullets/loads that are jumped, so seating bullets into the lands can certainly deliver excellent precision. That might be an option worth pursuing given the seating depth issue you described.
Finally, I understand that SAAMI lists minimum/maximum COAL values for .222 Rem of 2.040"/2.130" (p. 66;
https://saami.org/wp-content/uploads/2023/11/ANSI-SAAMI-Z299.4-CFR-Approved-2015-12-14-Posting-Copy.pdf). But bullet dimensions are critical to those measurements, as well as freebore length. So is there really such a thing as "minimum" COAL? Perhaps by strict technical definition, the answer is yes. However, there is almost certainly a lot more latitude in practice than those measurements might indicate. In the past, I've seated a fairly long, heavy bullet so far down in the neck of a .223 Rem case that the specific point on the bullet ogive that first contacted the rifling was actually buried well
below the case mouth. That particular rifle had a very short freebore...zero, actually. Even though I wouldn't ordinarily recommend doing this unless absolutely necessary, in that case it worked extremely well, and the rifle/load was capable of excellent precision.
In any event, it seems like the simplest solution to your issue without going to any further length such as throating out the chamber (increasing freebore) would simply be to carry out a seating depth test within the range that falls within the SAAMI specs. If you find a seating depth within that range that shoots well, problem solved. It may be necessary to adjust charge weight downward slightly if pressure starts to increase excessively as the bullet is seated into the lands, but that's not a big deal. Frankly, the easiest thing is just to start the entire test, both into the lands and off the lands, using a slightly reduced load for safety reasons. If any apparent seating depth optima are identified, either in the lands or off of the lands, one can then go back and re-visit charge weight if necessary. Best of luck with it, I hope you can get it figured out to your satisfaction.