Not enough data.
At +1SD (so an IQ of 115), you're sort of borderline for whether acceleration should be considered, even if all other factors are in favor, going by the Iowa scale.
+2SD (so an IQ of 130) is right in the middle of what some people consider "optimal IQ," so likely will do fine with no acceleration unless there's some other issue going on.
+4SD is generally the ceiling of the test. So unless you're going to use extended norming, or an outdated test, or in some other way fiddle the numbers, you don't have very much spread in the numbers to run the statistics. Dottie or someone could probably tell you the SEM for a score of 145 (+3SD) - I'd guess there'd be enough noise in the numbers that trying to draw distinctions between a kid with a 145 and a kid with a 155 based solely on one test would not be good science.
Plus, there really aren't that many kids who are skipped a grade - much less multiple grades. And lots of reasons other than IQ for a given kid to forgo a multiple-grade skip (or any skip!).
Ideally, you'd have a study whereby kids are randomly assigned to skip zero, one, or more grades, and the results are compared to those kids' IQ scores, but I'm not sure you could find an Institutional Review Board to approve it. Or a pool of parents willing to have their kids randomly assigned. Lacking that, you're unlikely to get any statistically meaningful data. And even given that, I'd think you'd be ethically obligated to try and help each kid succeed in the grade placement they got, and that would confound your results, too.
