It sounds like more of an issue of having a scale that isn't very precise, which is typical of a home scale. We just tried to purchase one to track the weight of our sick old cat and all spent 24 hours weighing ourselves a LOT since we've never had a scale. For the adults, it would vary weight up to 10# based on time of day, when we'd last ate, used the toilet, etc. For the 2-year old (about 23#), almost 2 # variance. For the sick cat (about 8#), 4# variance depending on how cooperative he was. Needlesss to say, I don't believe anyone's weight changed that much in 24 hours. Scale went back. But I can easily see how weighing a baby before/after a poop, nursing session, even just being squirmy, could mean a # weight difference, especially if the sensitivity is only 1/2# increments. Consider if baby was just below that 1/2" a month ago, had just nursed, and hadn't pooped that day. . . then a month later, baby hadn't nursed recently & had a big poop. . . easily could appear that baby weighed "less".
Personally, as long as your pediatrician is not over-reactive, I would set up regular weight checks at the Dr's office on the same scale each time. Perhaps weekly until you are sure she is continuing to gain, then monthly to track. If your ped is very reactive, though, you might be setting your baby up to for lot of testing and interventions, pressure to wean or supplement, that may not be indicated. Often IBCLCs have scales available and some baby stores do as well. Again, make sure to use the same scale each time and be consistent about time of day, feeding, and the like.