If you used CJK before UTF-8 was popular, you would know that people had absolutely no issue counting some characters as 2. Itβs entirely natural. In typical proportional fonts CJK characters still occupy about twice the width of a Latin character.
CJK users also don't use CJK characters exclusively. There's almost always single-byte characters mixed in, such as URLs and numbers, and in the case of Korean, even punctuation. If some characters count as 1 and others count as 2, it will result in fluctuating limits that are hard to grasp intuitively.