Well if any baker can tell me, then this baker just asks himself: how exactly is it more accurate to weigh the flour?
I've never understood this. Flour will gain weight when it has absorbed moisture, which means that your x grams of flour on the scale is really slightly LESS flour than you think. Or to put it another way, the total water content of your recipe has already been partially supplied by the atmosphere. Unless your weighing operation tells you how much water to withold or how much flour to add I honestly don't see how it helps to weigh it. Perhaps you know exactly how much additional weight you need for a given hygrometer reading?
Conversely, unless you know how much the volume changes between 1 dry cup and 1 slightly damp cup of flour, you can't do much with the volume measure method to account for moisture either. But I suspect that the changes in weight are greater than the changes in volume for a given rise in humidity, which means that weighing would be less accurate than measuring. Maybe I'm missing something, though. Is it that by also weighing the salt, you can be sure that the proportion of salt to flour is constant? Seems like salt would absorb a greater proportion of water per gram than flour though...
In practice, when I bake bread at home I know that I will use roughly three parts flour for every 1 part water, but I mix it in gradually and add or subtract according to texture. I previously worked as a baker both in a restaurant and in a small bread bakery. In the latter, bread was made by dumping an entire x lb bag into Hobart floor mixer, then adding y gallons of water, using a gallon pitcher. For quantities that large the effect of the atmosphere was negligible.
Disagreements among cooks aside, my main point is that cooking with the metric system on a small scale requires one to operate between the main units. The imperial system on the other hand, directly correlates between units and actions. 1 teaspoon means that the cook takes the teaspoon and fills it up once. The analogous measure by the metric cook is either to take a 1 ml spoon five times (hardly likely), or a 5 ml spoon once, which strikes me as a surrogate for a teaspoon. If they called 5 ml a metric teaspoon I'd be fine with that. Likewise if you're constantly referring to a 200 ml or a 240 ml measuring cup, you're using a surrogate for an imperial cup rather than making optimal use of a ten-based system. But measuring everything out in odd amounts nearly equal to imperial units seems to be using the metric system with a wink.
The ideal kitchen measuring system would be ten-based like the metric system but would have units that agreed with the impositions of nature (e.g. the size of an egg) and the practicalities of living. For example the basic unit for a small spoon-like measure would ideally be the amount of salt you'd use in a common dish, something like that. The unit of liquid measure should be the amount of liquid most commonly used such as the amount of water in a loaf. And the conversion of units should be in an easy-to-calculate system such as tens. IMO the imperial system satisfies the practical criteria better than metric, and the metric satisfies the conversion criteria better.
For other purposes the elegance of the metric system is undeniable. e.g. 1 mL of water at STP has a mass of 1 gram. 1 cubic meter of water is one metric ton. And the fundamental unit of lenth, the meter, while officially described as the wavelength of a a certain color of light (sadly, not Kentucky blue), is about the length of a person's stride.