I'm referring to this question: Natural linewidth of hyperfine levels?
Let's take Rubidium-85 for example. It's written everywhere that the D2 line has a linewidth of around 6 MHz. But the D2 "line", a transition between fine structure levels, is basically just a composition of the underlying hyperfine lines, which are shifted in frequency by tens of MHz relative to the D2 line.
How can the D2 line have a linewidth of 6 MHz if the underlying hyperfine lines are shifted by way more than 6 MHz? My only explanation would be that the 6 MHz already refers to the hyperfine lines. But this would mean that they all have the same linewidth?
What's the truth here?