The longitudinal stress adds to the failure mode, as far as I understand from the calculations that "technical papers" on the subject of stress indicate. But it is complicated because the tensile stress is combined with a perpendicular compressive stress (as the sum of squares so they become additive) but the stress limit varies depending on the stress of each vector. So to judge them independently is really the only way I know how, unless a mathematician can teach me otherwise.
I agree that safety factors exist. What I am saying suggests some old designs achieve safety factors of as low as 2 or 3 ish instead of 8, when compared to modern stress limits? Probably why "old" designs have not failed, but "not adequate" for modern thinking for insurance, certification, public exposure, personal safety, etc? (IMHO). It does not change the real "safety" of a boiler, as the metal does not change, but it changes the "risk" of failure (and litigation) when a boiler is de-rated to have a higher safety factor.
ASME seems very specific as to the limited tensile stress on pressurised copper boiler components, but I have not yet seen how it manages compressive stresses.
Hadrian's wall, the Colluseum in Rome and other buildings were not made to modern regs, but have lasted thousands of years. (Except for people robbing stones!).
I am just advising caution, and doing full calculations for any NEW designs, to the regulations appropriate to where you live/insure your stuff. On new designs, the calculations shall be checked by an appropriate expert as part of the initial certification (in UK practice). Designs that have previously been certified and proven are accepted on the basis of precedent. But I suggest you should be sure you are using good quality materials to suit those designs.
A bit like aircraft, this may suggest you could copy the design of De Havilland Comet window in a copy of that design, but the history records this design has been rejected following failures. So check that you copy a GOOD design with GOOD history, and without later changes being incorporated.
My Simplex boiler (after repair) will need re-certifying when I have made repairs (if the repairs are good!) But I shall de-rate it to 80psi NWP instead of the 100 psi on the drawing, as some flue tubes are "inadequate" for 100 psi compared to my interpretation of ASME Engineering ideology. Just because this may become a working boiler in a public area, covered by the Club's insurance, and it seems sensible to be as safe as "Modern thinking" requires for new stuff. Who am I to ignore stress limits or whatever determined by the experts in ASME? They have knowledge, expertise and history that didn't exist when Martin Evans and others designed their boilers.
So you do what you think is right and I shall do what I think is right, especially when advising people on a public site like this one. And surely as "Model Engineers" we should responsibly use the "best" within our knowledge, skills and expertise?
And I have not yet mentioned de-rating to account for distortion factors, stress concentraton factors etc..
There are plenty of worms in this can!
In my professional employ I had to consider far more than just the simple Hoop stress on tubes.
Please forgive my ramblings on this subject. Tell me to "Shut up!" If apropriate. No offence will be taken.