For me, billon is an alloy of silver (or much more rarely, gold) with copper/bronze, in which the precious metal content is low. If the coin is magnetic, then it's not billon in my book, however much silver it's got in it.
The imprecision of the term billon lies in the fact that the point at which coins stop being "0.xxx silver" and start being "billon" is very dependent on all sorts of things other than the actual fineness of the silver. Those big Mexican 1 peso coins from the 1960s are listed in KM as being "0.100 silver" (IIRC) but there are plenty of coins from the German and Italian states, and pre-Union Scotland, just described as billon. As far as I can see, the collective decision of the coin-collecting world to define a given issue as "billon" is down mostly to (a) convention (the turners and bawbees of the 17th century have always been called "billon" as far as I can tell), (b) lack of knowledge of what the proportions actually are or should be, and (c) related to (b), a lack of interest on the part of the original coiners in exactly how much silver was involved, because whatever it was, it was only a token amount and nowhere near the intrinsic quantity required to match the face value.
Conversely, where there is supposed to be a link between content and face value, the coins are generally termed "silver", even if the proportion of silver is very low. Some of the most debased products of Henry VIII's mismanagement probably had less silver than some Scottish small change or minor coins of the German states, but, debased though they were, they still had a precise amount of silver in them and are therefore referred to as "0.xxx silver".
In that context, I don't see that "billon" is any more imprecise than "cupro-nickel". In both cases they're mostly copper with "some" silver or nickel, where "some" varies and is rarely important for identification.