No one in academic circles doubt that there are multiple types of intelligences; how to break intelligence down into specific subdivisions is the current item of debate. For example, the Weschler Adult Intelligence Scale (WAIS)(the most common IQ test), which has been criticized for focusing too much on 2-3 specific types of intelligence, has 20+ subscores. A researcher can divide it up almost any way he wants.

Digression aside, I think what you're refering to is Howard Gardner's theory of multiple intelligences. He proposed that there are 7 types of intelligence: musical, kinesthetic (control over ones body, e.g. ballet dancer), spatial (mental representation of physical objects, e.g. maps), linguistic/verbal, logical/mathematical, intrapersonal (understanding of self), interpersonal (understanding of others, e.g. people-smarts, as you put it).

One way to find support for a distinct type of intelligence (and this is where neuropsychology comes in), is finding an individual with brain damage that impacts their ability in one area but not others. For example, if an individual were to exhibit brain damage that caused a huge deficit in their linguistic/verbal abilities compared to their premorbid functioning, this would support the idea of a separate linguistic/verbal intelligence. (There is: see individuals with damage to Broca and Wernicke's areas.)

Now that we have our definitions down, to answer your question: Gardner's theory has not been entirely discredited in academia, though criticism exists.
Hit up Wikipedia for Theory of multiple intelligence for a decent introductory description.

I could go on at great length on this topic, and get into areas that would be extremely boring to anyone not in this field, but I need to go to bed.
-Jon