Reliance on a CM for testing in a medium-low volume process?

[Pack34]

To define quantity, this is maybe 400-500 eaches a month.

To define the situation, the CM has test fixtures that they validate that the PCBAs are assembled properly to some arbitrary standard (the CM designed the fixtures). During final assembly, everything is assembled into the product and functionally tested. If some issue occurs the technicians on the line track down the suspect part, replace, and send it back to the CM for review. After this, they'll rework the board if they find an assembly issue and if not, the board is scrapped.

I can't help but feel that the process here is missing steps. All the information that I'm really getting is that there was an issue with a board and some high-level description of the potential failure. Such as: Board didn't power-up, Board powered-up but was unresponsive, Board powered-up and was responsive but didn't output the correct signals.

Shouldn't the process flow something like this?
1. PCBA flagged as suspect on the line
2. PCBA replaced and product shipped
3. PCBA disposed to engineering for review
3a. Thorough visual inspection of part
3b. Replication of failure mode using a functional test fixture
3c. Low-level test fixture analysis (meaning the board is flashed with a test firmware to throttle all I/O and power cycled)
3d. Issue is isolated, reported, and a determination is made to queue for a design improvement if needed


Now, I'm coming from essentially "starving start-ups" and still adjusting to larger volume processes, but I'm having difficulty advising on the disposition of materials if I don't have access to the parts nor any thorough FA notes.


What's the typical process seen here at this level of volume? Is it typical to offload so much to the CM? It feels that vital information is lost by an over-reliance on the CM that could direct robustness and reliability improvements.