I think it is only possible to speak generally about potential reasons for the material deformation. Either, or both, residual stress from welding (or forming) and thermal stress in service can induce structural deformation. My subjective opinion, without more service related data, is that given the duration of the service interval, the in-service thermal stress would more likely be the controlling cause for the deformation. Moreso if the component experienced regular thermal cycling in service. Too, a residual stress bias may superimpose with the service stresses to cause a resultant deformation.

It would be interesting to see what a post weld solution treatment of the component would accomplish. Once solution treated, a final shape for critical geometric surfaces could be machined and there would be higher confidence the geometry would be maintained in service.

As an example of what differential heating and cooling can do to a stainless steel structure, people where I used to work would indulge in "government" work (work by and for the people) and build themselves [austenitic] stainless steel mufflers for their cars. These would eventually shrink inward such that the internal baffle support structure could be seen in relief. The effect looked similar to the protruding ribs of a starving person. One muffler eventually tore itself open.

Without significant, internal or external, pressure, what could cause the observed deformation of the mufflers? The only reasonable force would be the differential thermal stress due to the differential thermal gradients cyclically propagating back and forth through the structure.