It's not wrong, it's close enough. And the point it works with more numbers and more type of calculation. Let's calculate 4% of 1243. That's the same as 1243% of 4, right, much easier to calculate by simply changing the 2 numbers... While my method is the same, by simply rounding everything.

And in engineering you always multiply/divide your results by a 1.5 or 1.25 safety factor, depending on situation. So you don't have to calculate exact results, just close enough. E.g. G is always 10m/s2. π is only 3.14, the other digits doesn't matter.