Knowing how much effort goes into commissioning a new variety of RAM chip I can tell you that, however good your soldering is, there's a non-zero chance that if you change the RAM you will see random crashes, most likely under heavy loads.
Swapping an identical (same part-code) chip for repair would work if you are sufficiently skilled. However, due to AI-related chip shortage density, org and other tuning/timing params are built in because we have to support hugely more SDRAM SKUs these days.
Where are you getting all these dead CM5s from ?
Always interested in innovative audio startups needing help and investment. Look for InPoSe Ltd or Future Horizons on LinkedIn to find me (same avatar photograph)
We buy RAM in large quantities on very competitive commercial terms. However, there may be occasions where some parts become available cheaper from third parties. These are likely to be in small volumes and completely untested by us. If labour is cheap enough then there may be a minor financial gain in buying a part with say a 2GB part, swapping it for an 8GB part of dubious origin, and reselling it as an 8GB device. However, this brings an element of risk to the user - the board hasn't been tested by us - and is a potential support burden for us, with customers complaining of unreliable devices. We therefore remove the commercial incentive by locking devices to their original RAM size. And for improved compatibility with a variety of suppliers, we now also program in some other device attributes, meaning that switching parts of the same size may not work. I don't see RAM being swapped between authorised devices of different sizes as a likely use case, but don't waste your time trying - it won't work.