Re: Thickness *LINK*

Maximum operating pressure is determined by the thickness and strength of the shell material. Additionally the effeciency of the joint must be taken into account. On a new shell or one of consistent thickness, the joint area usually determines the pressure since it is weaker than the shell because of the rivet holes.
#9 has a lap seam which is the poorest joint because its geometry forces the shell out of round. When pressure is applied, the joint has a tendency to flex as the shell tries to become perfectly round. By the early part of the last century they had figured this out and quit using lap seams after numerous failures. Sometimes, as in the case of #9 a reinforcing plate was applied over the joint to stiffen it and take some of the stress. It doesn't fix the problem but it does mitigate it somewhat. By the rivet pattern on #9 it appears to be a reasonable design and probably will give decent efficiency.
Railroads and the ICC were most concerned with the longitudinal joints on boilers and less concerned with wastage and the firebox area. Fireboxes and staybolts were items that usually got replaced because they wore out from use and cracked or leaked causing operational problems. Today we see corrosion and sheet thinning to a far greater extent than the railroads did simply because the locomotives are not used and are often stored in less than ideal conditions.
Any restoration must consider the type of service the locomotive is to see. A boiler with wastage on the sheets can be perfectly OK for service of a limited nature. If the service is to be severe, the wastage can lead to earlier failure.
It has been my experience that the railroads and builders did quite a bit of experimenting in their quest for better boilers. Some of the experiments worked. It seems that the boilers became fairly consistent after 1920 but the boilers built before 1910 seem to have more surprises.
Most of the calculations used for steam locomotives were derived from practical experience. They are not exactly engineering calculations, rather they are numeric litmus tests that indicate whether the component will function in a satisfactory manner. Today it would be possible to generate a model of the shell that could be used to determine the exact strength of the entire unit. This is currently being done in modern boiler construction. The problem with steam locomotives is that they bounce down the track and are subject to many stresses and strains that are unpredictable. An accurate model would have to take all of these into consideration. Think ACELA.
In concluding this long-winded post, if the boiler is surveyed by experienced people who intuitively know steam locomotives, and the calculations are done in the context of their derivation, a pretty accurate picture of the boiler should be the result.