OK, so I've been using the chart on wikipedia. At first I thought the chart was incredibly generous, then thought it was pretty accurate and now think that it may even be harsher on the ACT than it should be.
Basically, I got a 32, so that is really the only score I have done any sort of analysis of.
So scores 32 -->36 are all on the 99th percentile.
According to Collegeboard, here
score on the 99th percentile is a 2200. Obviously, it goes without saying that that is on the higher side of the score ranges on the wikipedia chart (2160-2210).
Next I went to princetonreview.com and looked at the published SAT 25-75% scores to see if they were consistently concurrent with the 25-75% for ACT scores. I went looking for schools with 25-75% scores between 28-32, so I wouldn't have to extrapolate a 27-31 or something to a 32; that seems like it could skew the data a little bit. Johns Hopkins
: SAT range: 1960-2240, ACT Range: 28-32...does this not mean that to Johns Hopkins admissions people a score of 32 on the ACT is equal to around a 2240? Wesleyan University
: SAT range: 1980-2230, ACT Range: 28-32.........does a 32 ACT=2240 SAT? University of Southern California
: SAT range: 1940-2190, ACT range: 28-32.........32 ACT=2190 SAT? Cornell University
:SAT range: 1950-2230, ACT Range: 28-32..........32 ACT=2230 SAT? Massachusetts Institute of Technology
: SAT range: 2100-2330, ACT range:30-34.........32 ACT=2215 SAT?
...and there are many, many more. Even at schools were the ACT is presumably the more prevalent of the tests.
University of Michigan: SAT range: 1850-2140, ACT range: 27-31.....32 on the ACT= (2140/31) *32 ==2209?? Again, that process probably leads to inaccuracy, but it's the best I know how to do. Tulane University
: SAT range: 1890-2165, ACT range: 27-31......32 ACT==2234??
University of Wisconsin: SAT range: 1790-2080, ACT range: 26-30......32 is equal to 2218??
(Probably some better examples for this category, but I'm tired and I'm sure the point has been proven as well as I can prove it)
Anyway, it seems as if all the SAT/ACT concurrence data is based off of the Collegeboard table that was compiled from information gathered between 1994 and 1996. That was over a decade ago and the test has changed significantly since then. Not only that, but wouldn't collegeboard have an agenda in getting people to think they did better on the SAT than they actually did in order to get them to send more score reports for $10 a piece? That's probably a bit beyond the realm of believability, but I digress.
The best I can figure is that a 32 is equivalent to at least a 2210 or so. It could just be wishful thinking, but the stats seem to be there to back this up. If not, please tell me where and why I'm wrong.