Honestly the lab kits you buy online can be all over the place and a lot of them aren't calibrated the same way the city's testing is. Utilities test at the plant and at points in the system, not at your kitchen faucet, so of course the numbers won't match up exactly, that doesn't mean they're hiding something. Also a $30 strip test is not the same as a certified lab test, those things can be off by a lot depending on how you use them. Phoenix water is known for being hard and having high TDS, that's not a secret, it's on their yearly report. If you really want to know, send a sample to a real lab that's certified, then compare that to the utility report, otherwise you're kinda comparing apples to oranges.

How many hours of stress testing are you actually putting on those refurbs before they leave the shop, or are you trusting the SMART data and calling it good? The thing that would worry me about using one for an emergency swap on a machine with no backup is that you just moved the client's only copy of their data onto a drive with unknown history. What's your plan if that refurb dies in three months and they come back at you? SMART can look clean right up until the moment it doesn't, and ddrescue doesn't tell you anything about how much life is left in the platters. If you're going to make this a regular thing I would want at least a full surface scan and a week of power-on hours logged before it touches a client's only drive.