There are a number of "we've always done it that way" things that don't really fit with the Code.
On the other hand, a lot of the codes have been developed from "we've always done it that way," if "that way" has proven itself to be safe and reliable.
When's the last time any of you saw a failure at an HVAC unit related in any way whatsoever to NM running through LFNC? I can buy the argument that watertight conduit underground is considered to be immersed with a high risk of filling with water and must be treated wet, but watertight boxes are considering dry locations, and homes with siding are considered dry locations, so why not the interior of watertight conduit connecting those two dry locations? If we consider that conduit to be "wet", then what about the "dry" boxes it connects to that all that water will be pouring into? We have to draw the line somewhere. If the code says it's wet through some technicalities, that doesn't necessarily mean it's wet; the code may need revised instead.
Likewise, I can't recall the code offhand, but isn't cable fill N/A or very relaxed for nipples and short runs?
