Ethernet has been standardised for over 30 years. But you are right, I probably have forgotten more about networks than you will ever know.
Whilst I doubt that's true, it's more about what we both currently remember than what you've forgotten, isn't it? (Yes, I know you weren't being literal, just haughty).
To date, from implementation of (and including) IEEE 802.3 in 1983 there have been 51 amendments accepted - the most recent late last year and 3 future ones already proposed. This excludes the experimental Ethernet in 1973 and DIX v2.0 in 1982. So standardised, absolutely, but hardly static. And, more importantly, use of that and implementations in operating systems and drivers and devices and hardware have all changed and matured.
There's really no reason to usher people away from using their printers over a network. There are more advantages than disadvantages and the fault description, along with the known background that the printer is in communication with the driver and related software (ink levels), suggests a problem elsewhere. One that's common, and quite easy to fix most of the time with a little guidance.
Yes, testing with a USB is a good idea and well worth doing and if that had been the sum of your advice I would have agreed whole heartedly. But suggesting that networks are something to be avoided in this case ("if you must use...") is just nonsense, sorry. I'm glad USB meets your requirements, of course, and for others that will also be true. But there's no need to avoid either method and certainly no need to point to networks as being particularly difficult things in a typical setup that photogs would be using in small business or at home.