I recently got a detailed cognitive assessment back and the subtest level scores were way more interesting than the overall number. My fluid reasoning and processing speed were quite different from each other and the composite score kind of averaged them into something that felt less true than either individual score. Is there a legitimate argument that reporting a single IQ number is actually misleading and that the profile underneath it deserves more attention?
Your intuition is right and a lot of practitioners agree with you. The full scale score is useful as a quick summary of general cognitive ability and g is real and predictive, but collapsing a rich cognitive profile into one number inevitably loses information that matters. The subtest scatter you noticed, where fluid reasoning and processing speed diverge significantly, is actually clinically meaningful and can point toward specific learning profiles, attentional differences, or processing characteristics that the composite completely obscures. This is exactly why the CHC model separates broad abilities into distinct indices rather than just reporting one number 
The single number is like a restaurant rating out of ten 
Useful for a quick comparison but completely useless if you want to know whether the food is good and the service is terrible or vice versa. The profile underneath is where the actually interesting information lives 