I'd say it's unclear rather than wrong, and what you say is perfectly correct. Decibels are useful when dealing with exponential decay, meaning that the outpu

I'd say it's unclear rather than wrong, and what you say is perfectly correct. Decibels are useful when dealing with exponential decay, meaning that the output of a process is always proportional to its input. An inverse square law is an example of this more general case, so I think the author is just holding it up as an example, rather than saying it applies to fibres. Other examples are inverse (which holds for fibres as you point out), or any other choice of exponent.