Interesting observation, Paul. I have to think, though, that whatever you observed was not due to density. The density of water changes from 1.000 g/mL at 4-C (the way the gram was defined in the metric system to begin with) to 0.998 g/mL at 20-C. Those are at the lower and upper ends of the temps we typically fish (40-F to 70-F), respectively.
Now, water's vapor pressure does change more significantly over that temperature range (from 6.1 mmHg to 17.5 mmHg), which might be a better indicator of how much tension there is between the molecules at those temps (vapor pressure, for those who don't know, measures how much pressure a liquid exerts on its surroundings, which means it is inversely related to how well the molecules are sticking together as their kinetic energy increases with temperature). Surface tension, meanwhile, drops from about 7.49 N/m to about 7.28 N/m over that same temperature range, a small but perhaps noticeable change. I'm sure there are other measures in fluid mechanics (the difference in the way the water shears around a boundary such as flyline at different temps, for example) that might be more useful to a flyfisherman, but I don't know what they are. Anyone here have a degree in fluid mechanics?
Water temp has a large effect on many things, but density is not one of them (its density changes by only 3% from melting to boiling point). It's possible that something of significance to the drift of a fishing line changes significantly with temperature, or perhaps other factors between the two water sources you were comparing caused you to observe effects in the water that seem more compelling on an anecdotal level than on a scientifically measurable one, but I'm almost certain that density has little to do with it. That said, regardless of what to call the effect, if it matters in terms of the drift of the line in the water, then it matters.
I'd love to hear what some of the more experienced fisherman than I have to say about this.
-Shawn