Really interesting stuff. Thank you for posting, Rajat!
I definitely believed that schools wouldn't care about the IR section for at least another three or four years, but maybe the validity studies will prove IR's value far earlier than I thought. I actually agree that the IR section seems to test more important skills than the rest of the GMAT (data interpretation is far more valuable than number properties or modifier placement, right?), but I figured that b-schools already have plenty of data points on applicants, and the IR data point is unlikely to say anything terribly new about any given candidate.
And it looks like I'll be very wrong about that. Rajat, you're completely correct: if the next batch of studies proves that somebody with a 700 composite and a 5 on the IR performs worse in b-school than a student with a 700 composite and a 7 or 8 on IR, the game will change really quickly.
For whatever it's worth, I'm a little bit skeptical of this first batch of data. Despite GMAC's best marketing efforts, many test-takers aren't taking IR seriously yet, and some well-prepared test-takers (including many of our friends in this forum) are likely to conserve energy on IR, and therefore get a lower IR score relative to their quant and verbal scores. Maybe it's not a large effect, but I wouldn't be surprised if the correlation figures are skewed by the fact that many test-takers refuse to take the IR seriously yet. My bet is that the correlation between IR and quant/verbal scores will slowly inch upward, but not nearly enough to invalidate the IR section.
It will be really interesting to see what happens when GMAC has more complete data over the next year or two. Despite my mild skepticism about this first batch of data, it looks like GMAC is well on its way to proving that the IR section matters, and soon it will just be a question of how long it will take for adcoms to fully embrace the data.