It could be more accurate to say that pi=3.1415... is wrong. Euler seemed to use the pi symbol the way we now use the theta symbol. So sticking to the radius may just be stubbornness.