I'm not sure I have much to contribute, because I haven't had the problem. But will inject my 2 cents, which may not be worth even that much.
Our old Pace Arrow had a simple ~2 inch fill port with a ~1/4" plastic tube coming back from the tank top and mounted at the top of the fill port fitting that vented things. When water came out that tube or backed up the fill port,you knew it was full. You often had to wait a minute for water and air to stop blowing out that tube before plugging it and capping the fill port. But simple it was.
There is no obvious vent tube end in my water bay, so I am presuming that rather than a separate vent on the tank, air escapes via the overflow valve until the tank is full, when water escapes there. Now I don't know exactly where the pressure regulator is, but I assume it is behind the fill port fitting - many of you that have replaced such an animal certainly aren't as ignorant as I. It may be where it only affects the "city water" connection pressure, on that side of the switch, rather than where it controls the tank fill rate as well.
Regardless, if the input pressure is higher than the overflow valve's rate capacity, pressure will build and the tank flex, resulting in some residual overflow of water after the faucet is turned off. But the term "siphoning" here doesn't quite seem correct, as there is no lower pressure being created outside, as much as there is residual high pressure in the tank. I guess it depends on how you look at it... lower outside ambient pressure "sucking" or higher tank air pressure pushing. So for 25 gallons to be pushed out, it occurs to me that the overflow valve is unduly restricted, the tank walls are enormously too flexible, and a tremendous amount of air is trapped when the faucet turns off. Or the same results occur because I am wrong and there is a secondary relief vent valve that's malfunctioning. If the regulator behind the fill valve is faulty or non-existent, one could be allowing far more pressure from some sources (60+psi) than the venting system and tank wall flex can handle.
I can easily see my translucent tank since it's up front in the bay with the HydroHot. When it fills to the top and overflows because I got distracted during the long fill and wasn't on top of things, I can see that when the overflow stops there is little air space at the tank top. So nothing is really "siphoned" out, certainly not a quarter tank. The water gushing out of the overflow seems at a rate little more than the fill hose end, and stops within a minute. A lot of this I'd presume goes to just how flexible the tank is, and that may vary from coach to coach, I dunno.
If there was an auto-stop on the fill, the overflow valve would seem of little use except as an air vent. On ours, the overfill valve would never stop if I let it and was that way when it was new, so I know we have no auto-stop on the fill.
The pex line Dick sees could conceivably be a roof AC drain, although ones I've seen are simple poly tubes and route down an outside wall all the way, not turn toward the center of the coach. But perhaps techs just took the easiest way down and out via a hole already cut for other lines. And I've seen online vent valves or vacuum breaks that one could add directly to the tank's top if its accessible and there's room; that would provide an exit for air pressure beyond what the overflow handles. But getting to the tank top and cutting a hole for a valve may be not be feasible.
Forgive my ruminating here as I know it comes from some ignorance on how our Beaver water systems are configured; the Pace Arrow's was far more accessible and simple. But perhaps I've offered some thoughts worth considering.
-Joel
God Speed, John Glenn