It  is something on the verge of a fairy tale to think that white smoke 
 would necessarily signal a coolant leak to the cylinders. NOrmal combustion products will
 yield 1.3 gallon of water (exiting as vapour ) for each gallon of fuel burned and this
water  condenses in the exhaust system and drips down until the pipe is hot 
enough to carry the vapour as such to the tail of the exhaust system.
fuel  +  air 
C7H16 + 11O2 + 44 N2 = 7CO2 + 8H2O + 44 N2
by molecular weight:
100 + 352 + 44 N2= 308 + 144 + 44 N2 (452 + 44 N2 on both sides, nothing gained, nothing lost) 
thus 1.00 pound of fuel C7H16 requires 3.52 pound of O2 (or 15.179 pound of air, the rest
 being Nitrogen carried through almost unchanged) and yields roughly 3 pounds of CO2 and 1.4 
pound of water as combustion by products.

at 60 mph, when your car burns 3 gallons per hour, you have like 4 gallons of water coming
 out as vapour....a lot of vapour...so even emptying the entire cooling system in one hour 
would not generate as much "white smoke" as what already goes out normally.

The only case where this near urban legend holds true is when the leak occurs at rest and water 
accumulates in cylinder or exhaust manifold and on a sudden, at start-up, all of it suddenly 
evaporates and would procure a heavy white cloud, really out of the ordinary, and this, for a
 limited time. But a slow leak that occurs only when engine runs is not detectable at the exhaust.