Unnecessary to perform charpy, or spectrometry, or other destructive testing if you have material/process certs from AS or ISO suppliers or vendors. They aren't aging the aluminum to T6 condition, and even then, it's the HT facility that is responsible for this cert. NDT performed on aluminum components machined is also rare in aerospace/defense let alone a bush league manufacturing outfit (compared to AS shops) unless there are areas that run risk of cracking which should not be the case if the design is good and the machining process isn't ####ed up. We do FPI/XRay on welds, on chrome, HVOF, or cladded or similar coatings after grinding, but extremely rare to perform this testing after machining on aluminum. In fact after engineering hundreds of parts going on fighterjets or commercial landing gear, I haven't seen one that gets FPI'd because of machining. Hell these things aren't even shot peened, there's no need, the QA requirements aren't even in the same league let alone ballpark. If you mean things like the Bolt, then thats another story and these should be MPI tested but charpy isn't needed if following the proven HT methods using AS/ISO certified vendors.
With regards to in process inspection, it depends on the size of the outfit and the qty being machined. No shop is going to outfit themselves w optical measurement feeding SPC for a run of 600-800 receivers which is basically what it works out to from experience. You'll ideally be using gauging (snap, plug, ring, block, etc) and equipment that makes financial sense and has a level of risk mitigation built in. For Diemaco/Colt however, a true SPC gauging setup is ideal, however optical measurement fails often and is always backed up by analog measurement as alternate so there's little point in this sort of volume where PMM isn't a thing (its not automotive, vomit). When I was process engineering for a landing gear manufacturer we avoided as much CMM and optical measurement as possible as the repeatability for a lot of dimensions was poor compared to analog measurement where allowed (IE scanning a bore on an average CMM will result in differences of tenths using same program and not moving part, the accuracy of most CMMs are +/-.0002 which surprises people reading the product data vs actually running the equipment) so I wouldn't judge a shops lack of this equipment as the reason for QA escapes, but rather their process, QC strategy, and what D-Key dimensions they are inspecting to begin with.
Their failures have more to do with their design, and QA than it does lack of CMM or sending certified material out for testing. I also don't doubt they are using measuring equipment wrong, which for guys like us makes us cringe!