Some good articles on both sides, and I liked the article posted by dropzone. I do think there is some confusion here, however, that requires a little good philosophy to clear up.
Whether we convert or not, we still have the problem of standard sizes.
For historical reasons, the US has been built–industry and construction–on a set of standard sizes denominated in inches, pounds, etc. But the fact that they are so demoninated is in reality a superficial issue. Those sizes were chosen for a reason, or for no reason at all, and we are stuck with them. We should change those sizes IF the cost savings justify it, and we should keep them if they do not. And if we change those sizes, they should be changed to ideal sizes, which may or may not be denominated in “even” metric units (although, if so denominating them saves costs also, then that should be taken into consideration too).
The “meta” argument is that we should convert to base 12.
People talk as thought the metric system is ideal simply because it’s base ten. But we can one-up that argument by pointing out that base 10 is not ideal in the first place. Base 12 is clearly superior because it has more factors (12, 6, 4, 3, 2 instead of just 10, 5, 2) and has non-repeating, single-digit duodecimals for 1/3 (0.4) and 1/4 (0.3).
People used to argue for a conversion to base 12, but they don’t much any more. The costs would be just too great, both in terms of the economy and culture (every number in every book would become archaic).
But the same is true for converting rigorously to metric, though to a lesser extent. People have done so where it is profitable and useful, and have avoided it where it is not. That’s also why we still use QWERTY keyboards: the initial cost of converting to a better system is so great that the NPV of the whole project is negative, even when you look at the long term.
The metric system itself is old and clunky.
It is, after all, a system that was designed in the 18th century. It’s sole original virtue was the use of base ten and the relation of one unit to another (the fact that modern science has continued to build on it is another thing in its favor, but not really a virtue of the system itself). Whether the units themselves were good for what they were designed for was really a matter of luck, and in many cases the luck was bad, and the intentions of the designers unrealized.
For example, the gram was rejected as the main unit of weight because it’s too small; science went with the kilogram. Sometimes booze is labeled in cl, but for the most part the middling centi and deci units for macro items have remained unused. (Micro is another matter–see below.)
Keeping this in mind, metric fails where Imp./US standard succeeds in many cases. In this thread people have said that it’s all a matter of what system you are used to, that neither is more intuitive. I disagree. Feet continue to be used, but why are decimeters not? Simple: the decimeter is too small to be useful for big things, too big to be useful for small things, and there is no nice chunky fraction (1/4 or 1/3, say) of a meter available. My feeling is that people use cm for measuring, say, the height of people and whatnot simply because they are in a metric society and, although deci is in theory an option, cm is the lesser of the two evils. The kilogram, however, is chunky enough to be useful.
Chunkiness is important and explains why other almost wholly converted contries still use some old measurements. In Japan they still use tsubo (3.306 m[sup]2[/sup]) for land area because it relates well to plot sizes in Japanese cities. The gou and shou are still used for sake and shouchou volumes. Kimono silk is still sold by the tan. And if I’ve heard correctly, China still uses a ton of the old measurements alongside the metric units.
Kilometer has probably not overtaken the mile for a very simple reason: it’s a long, cold-sounding word, whereas mile is short, familiar, and sounds good in literature and poetry. Branding is important. The same goes for kilogram and pound.
If we were going to design a system of measurement today, I doubt we’d create the same units. It would probably be wise to create a parallel and compatible folk version of the system with some units not in base ten for greater usability. But, then again, base 10 has few factors, which brings us back to the problem noted above.
Metric is a hit in micro, a flop in macro.
Chunkiness and intuitiveness simply don’t matter with micro measurements, since we can’t see the stuff anyway. The miligram and micrometer (often branded “micron” for ease of saying) are hits, and the nanometer is simply great for what it does. Here the metric system reigns supreme. It’s a needed thing.
But metric simply isn’t needed for macro units, and never was. The reason is that the multipliers are useless. Once you have mile, ton, and gallon, you’re done and you don’t need another unit. Nor does anyone need megagram and gigameter: they simply aren’t used.
Precision at these levels isn’t required, and if you have to convert from metric to US, you simply convert. Standard sizes aren’t an issue. Nothing is. That’s why miles, pounds, and gallons are just fine for everyday use.
In conclusion
Metric has won where it does the most good: in the microworld and in complicated physics and engineering calculations. Elsewhere it doesn’t offer an advantage.
And framers, well, we won’t talk about them.