Erm, you're onto a very important discovery. Stephen Jay Gould (1981) discussed this notion in depth. Intelligence is not a single entity and its definition is at best vague and at worst incoherent. Its true that people who do well on one type of a cognitive task tend to do well on many others, however, this does not show that we know their innate intelligence. Many of the tests can be mastered by rote learning and focus on culture-specific phenomena and do not cover taks that assess important cognitive abilities. They also ignore a deep conceptual issue regarding the question of whether or not IQ itself is a coherent notion. Consider for example the work of a Mathematician and compare it to that of a historian. Both scholars would be required to proficiently use imagination and engage in deductive reasoning. However, mathematics requires more deductive reasoning than brainstorming and vice versa is the case with history. Clearly, some people are better at one than the other and some are more talented at one than the other. Even among mathematicians there is a rift between problem solvers and theory builders. It does not exactly correspond to the discrepancy between visionaries and deductive analysts, but it does point out that people who are clearly intelligent tend to excell at one type of an activity more than at the other. The foundational notion of IQ testing, namely that there is an exact figure assessing how smart a person is, or how well he will solve all problems is mythical. At best, there are vague approximations regarding how a person will fair in most intellectual affairs, but the lack of nuance greatly undermines the argument of the proponent of IQ testing.
You should, however, clarify your conceptual distinction between IQ and intelligence. I am guessing one means a person's IQ test-result and the other is intelligence. How is intelligence a scalar quality? The entity in itself does not yield to a precise measurement. It is an agreggate of multiple cognitive abilities. If IQ tests were able to test the functioning of all abilities relevant to problem solving the result would at best be an average of all of those scores, which would be symbolized as a single number. In itself the number would tell very little about how a person will solve all problems. For example, he could be better than 99.99% of contestants at memorizing items, yet inferior to 7% of other test-takers in using his imagination but also superior to 98% of his competitors in deductive reasoning. In order to know how well he will perform in a certain intellectual task, we'd be well advised to determine how well he performs in specific cognitive activities that are required by the task the most. If 'general intelligence' is not an incoherent concept resultant of a hasty generalization fallacy, tell me what is!
A challenger of the theory I have advanced above may claim that people who are good at analysis are also imaginative and have a formiddable memory. For the very least, in most cases, when a person's displays competence in one area, he can develop skills in others. Such an example of course excludes a number of idiosyncratic cases, such as the Savant syndrome.
There is a certain underpinning for intelligence, but it is rather vague and abstract. I cannot provide an accurate descriptive account of this phenomenon without relying on analogical reasoning and concrete examples. Lets imagine a man who is 'in shape', suppose that his regular exercising program consists of lifting weights, running and playing hockey. In all of these activities, he uses various skills that will prove timely in other physical contests that he may encounter. For instance, he would be more likely to succeed as a soccer or a tennis player than someone who is 'couch-potato'. The case is such because he regularly cultivates many athletical skills consistently, just as the historian, the mathematician and the political scientist exercises all cognitive skills. It is true that they rely on some more than others, but nonetheless, they could prevail in academic tasks that require a different type of cognitive work than the kind they're accustomed simply because they've practiced them more. Thus, in order to do one type of intellectual work, you need to participate in nearly all other kinds, at least occassionally. This, however, does not mean that you perform equally well in all of them and precisely for this reason, the concept of intelligence is rendered untenable. To assert that a person has a general intelligence is as grotesquely absurd as to claim that he also has AQ or general 'athleticism' the implication of which is that he is equally skilled at all athletic activities that we may imagine! This means that he is equally good at ice-skating and weight-lifting. The range of cognitive activities is nearly as wide as that of athletic abilities. As a soccer fan, I often notice that not all players are equally good at all activities that pertain to the sport of their professional specialization. For example, some are excellent shooters, yet inadequate tacklers, some have great pace, but encounter terrible difficulties shooting accurately on target. If we see such disparities in one sport, it would be hopeless non-sense to assert that a person's competence at one athletic activity can be generalized to all possible athletic activities. Similarly, a person who has the skill-set of a typical academic such as analytical reasoning, pattern recognition and memorization may be destitute of other important intellectual abilities that we often overlook. I can comment on this from personal experience. As a philosophy student I have often observed English, Historians and Art scholars struggle with philosophical concepts. Similarly, many students and professionals working in philosophy, physics or mathematics struggled as literature or art critics. Almost none of the analytic philosophers have the kind of a prose that mathces the artistic talents of the more artistically minded scholars. For this reason, academics of the exact sciences have alienated the humanities and vice versa. Its not that they simply have no interest in the disciplines of one another, but its a fact that the scholars are more talented at one academic enterprise than the other. They are often able to succeed in disciplines that are similar to their own, but not in those that are significantly different. For instance, many mathematicians have abandoned their discipline in favor of philosophy, yet fewer of them have become historians, art critics of English scholars.
At this point I have limited my discussion to strictly intelligence or ability to solve various complex problems. I've purposefully omitted any reference to the concept of 'innate intelligence' as it is far too vague and hopeless confused to be addressed in a serious conversation. To say the least of it, the authors of the Bell Curve claimed that the intelligence and IQ test-scores of children do not change throughout their lives. Famously, the American Psychological Association found no evidence to support their claim and stated that intelligence is mostly a result of a person's experience with the environment. (See Bloom 2007, Lecture 13
Session 13 - Why Are People Different?: Differences — Open Yale Courses)