I have recently been reading the likes of Owen Barfield, Rudolf Steiner and (excerpts from) Goethe on the nature of scientific thinking as it should be - and this has clarified for me that real science requires real understanding which requires bringing the subject to life in your mind so that your mind is actually thinking-it (thinking the entity, the object or more usually the process).
Therefore science requires understanding ,and understanding is qualitative not quantitative - understanding is not about description and measurement.
(These are possible when scientific understanding is utterly lacking; - they can even be done mechanically even, by a machine - although there must first be some-thing to draw-a-line-around that which is described or measured.)
Before science can be done, there must be this qualitative understanding of the phenomenon as an irreducible entity - not in terms of other things, but in terms of itself.
This kind of understanding is more lacking from intelligence research, even from the best of intelligence research, than in almost any field I have encountered (although Epidemiology runs it close) - indeed the lack of understanding of intelligence is often celebrated, and a 'theory-free' activity of intelligence description, measurement and correlation is put forward as - not merely an expedient or temporary necessity, but - the ideal.
People therefore misunderstand the IQ score as the reality, and intelligence as merely mathematically-derived from IQ scores.
But to understand intelligence requires understanding that a person may be of high intelligence and not have a high IQ score (no matter how validly, often and carefully the IQ is measured and calculated); and another person may have high IQ scores (measured in the best ways and by the best methods) yet not be of high intelligence ^.
This is a simple consequence of the fact that intelligence is not reducible to IQ score or any other measurement: it is an indivisible, qualitative entity.
Intelligence can be 'thought' hence understood - but understanding intelligence is like knowing a person (as compared with describing a person, their hair colour, height etc); getting to know intelligence is therefore like getting to know a person - it is a result of experience.
^ Highly intelligent people who do not score highly on IQ tests are easy to understand - because anything which reduces test performance could lead to this outcome: illness, pain, impaired consciousness and concentration from sleepiness, drugs, drug-withdrawal, mental illness... there are multiple causes, and some are chronic (long-lasting, perhaps life-long).
People with high IQ scores who are not of similarly high intelligence are familiar to anyone who has attended a highly-selective college or educational programme or who are members of intellectually 'elite' professions; since they make-up the majority of participants.
Note added: The 'Flynn effect' - of rising average IQ scores in a context of reducing average real-intelligence - is an historical record of the emergence of more and more people with high IQ scores who do not have similarly high real intelligence - until now it seems likely that many or most people with among high IQ scorers do not have similarly high real-intelligence - and this applies especially at highly-educationally-selective institutions where Endogenous personalities are substantially selected-out. The correlation between IQ score and 'g' was much higher in the past (a century plus ago) than it is now.