Not to be geometrically anal about it, but does measuring road distance give you an accurate rise-over-run? Specifically, the run part? If your hill forms a right triangle, then the short leg of the triangle is your rise, that's straightforward; but in a true slope calculation, the "run" is the bottom leg of the triangle, not the sloped leg (hypotenuse), which is what's being measured if you rely on your cyclocomputer's mileage. The difference is probably trivial, but might cause you to slightly underestimate the slope (too long a run for a given rise). Any math weenies have insight into this?