I think you're both right in some way. I would define efficiency as the amount of heat removed per unit volume of cooling water. By my definition, I think your view is wrong unless you replace the word 'efficiency' with 'speed'. I think your dad's opinion is close to the truth - lower flow rates will not work as well in terms of raw chilling speed, but they will work better in terms of efficiency (by my definition).
Of course higher flow rates will chill the fastest. Think about it from the wort's perspective. If the flow rate is low, only part of the chiller is cold - then it grows gradually warmer further along the tube. If the flow rate is very high, then the whole chiller is cold.
Heat transfer occurs proportional to the temperature differential, so the cooler the chiller is as a whole, the higher the heat transfer, period.
But, with an immersion chiller your biggest problem by far is keeping the wort evenly mixed so you don't have a cold spot around the chiller. At high temperatures, the huge temperature differential keeps the chilling efficiency pretty good even if you aren't stirring the wort, but as it cools down, the efficiency plummets if you're not stirring - you get a cold spot around the chiller and then you're removing VERY little heat. Even if you do stir, the efficiency is naturally still going to drop since the temperature differential is much lower.
Chilling water flow is a case of diminishing returns. When the wort is relatively cool, the temperature differential is low and running the water full blast will not keep the chiller all that much colder overall than if you only ran it at, say, 50%, so using more water is more or less a waste. So, most of us pay attention to the outlet water temp and throttle the water flow accordingly. If you throttle the flow down as chilling progresses so that it is always at least warm, you keep your efficiency up to an acceptable level, saving water.