ada4 seems to be the drive you should care about. You have a Multi_Zone_Error_Rate of 1. Others are fine, to me. Another note: you should not run a smart test every hour. short tests can be done once a week and a long test once a month or every 15 days.
Ah, bull. The whole point of SMART is to allow the early detection of a problem. Running a short once a week? Might as well not run it.
Short tests are deliberately intended not to interfere with the operation of a drive; every four or six hours works pretty well. Since this is the primary mechanism that exists to detect problems with a warm spare, and since running them at this frequency shouldn't impact a pool, this is quite reasonable.
The bigger problem is how to schedule longs. Longs are useful for identifying disks that are developing a problem, but they do require a full disk traversal of all LBA's. As with shorts, the design of SMART means that the drive does these tests in the background, but with a long test on a very busy pool, you could run into a problem, especially if a scrub or resilver was running at the same time. This was apparently discussed and dismissed as a nonissue;
https://bugs.pcbsd.org/issues/7362
For what it's worth, my take on the whole thing is that if your pool is SO busy that it cannot handle normal access, a scrub, and a SMART long test at the same time, you don't have sufficient headroom in your pool's design, and you need to fix your pool. We're running shorts every four (six?) hours, longs twice a week, and then scrubs at 35 days. This is an optimal design, because the hard drive is much better suited to handle identifying its own issues; unlike a scrub, a SMART long test does a linear traversal of LBA's, and does so without transferring the data to the host. It will locate issues fairly quickly, which is the key thing you want SMART to do.
Your strategy is bad because you might not learn about a developing media defect for an entire month.