The reason you never get that high is that it's not a pound of sugar plus a gallon of water that produces 1.046. It's a pound of sugar in a gallon of total wort. The wort volume includes the volume added by the sugar itself. In your example, the pound of sugar is being stretched beyond a gallon, so the gravity is less than 1.046.
That doesn't really work, though it's okay as an approximation. . As an example, let's take an OG of 1.068 (9% potential alcohol) and an FG of 1.008 (1% "potential alcohol"). Subtracting the potential alcohols, we'd compute 8% ABV. But applying a standard beer ABV formula, we'd get 7.88%. The reason the potential alcohol math doesn't work is that a potential alcohol reading tells you what ABV you'd get if you fermented a wort (or a wine must, really) to dryness. It's strictly speaking only useful for the "OG" reading. Once fermentation starts, there is alcohol in the mix, which skews the gravity downward (alcohol being less dense than water). Thus any "FG" potential alcohol reading implies ("as is," if not adjusted) less sugars (carbs) remaining than there really are.
I can't argue with anything you've said. I could indeed have been more precise about a pound in a total gallon. I'd just add that I don't think a floating hydrometer is a precision instrument anyway and then factor in reading through the meniscus, there's plenty of room for imprecision. OP is only looking for about a 1/2% increase in ABV. In a 5 gallon batch, maybe 8/10th of a pound of table sugar would get him in the ballpark.