divs and FCF approaches

Can someone please expand on this:

‘FCF approaches are best when dividends are paid but do not appear to be representative of the firm’s capacity to pay them.’

Aren’t FCF approaches best when do dividends are paid at all?

The first rule is a broader description of the second. Basically FCF is used when there are no dividends, but even if there are dividends but they are very irregular, again use FCF.

Normally questions specify the method to use, so this isnt the best place to get bogged down.