Not sure, but open to any/all ideas. I’m guessing the plastic is going to crack eventually as it gets brittle no matter how it’s reinforced- but not an engineer either. But I did stay at a holiday inn last night ;-)
There are plastics that can withstand harsh environments. Maybe they need to be fiberglass reinforced, or be a more expensive material. Plastics can have wildly different material properties. They can also have very different costs - not just of the material, but for the tooling and or process.
Haven't read the whole thread, but resin selection for 3D printing is still very limited. The few printers that offer good selection often have fairly poor dimensional accuracy. This is important for properly sealing without secondary operations (millling/grinding/etc.). For a highly (thermally) stressed part, I'd be concerned about the anisotropy of the material in the part.
That said, it may all work wonderfully. But beware of assuming the engineers were a bunch of bums and a simple 3D printed part can do better. As is often the case, they were probably forced to squeeze cost out of it leading to issues down the line and they did they best they could and proposed a less expensive design that they felt would work.
The part probably fails for one of two reasons (or a combination of both):
1) The part is failing in fatigue based on either heat cycles (thermal expansion/contraction) or
2) stresses brought upon by dissimilar thermal expansion rates between the part and the part it is attached to.
Embrittlement or other deterioration (resulting from hygroscopic properties, UV exposure, etc.) would make matters worse.
Do we have a close-up view of the failed part?