The empirical data only tells you that your mass result is variable. It's your theorising that suggest as a hypothesis that the lake's current volume is perturbing your measurement - which you then show to be consistent with the rest of your theoretical framework by experiment.

I wonder if the bounds are such that someone walking along the corridor will affect the measurement ("yo momma so fat she perturbs the expected measured mass of the Z-boson!"). What's the relative gravitational attraction of, say, Jupiter versus a nearby train.

A great insight.

I'd love to be the guy who decided I bet that low frequency monthly term is the tides, let's just regress it out shall we?

Yeah, surely someone got a free beer out of that one.

I wonder, once you've convinced yourself the idea isn't crazy, how you go about convincing someone else --- No, look at the tidal chart, and then look at this variation here. See? And then one day you've hit critical mass and the whole lab is looking at train tables, the passage of the moon, and how often yo' momma brings you lunch, trying to find correspondences.

Or maybe there was already a list of possible confabulations, in decending order, and they checked them off one by one. That would be less amusing, I think.