Notice also that Romney and Santorum are tied at 34 among white Republicans in OR; Romney's lead comes by winning Hispanics 75-12. And Hispanics are 10% of the Republican sample, which is probably high since they're only 9% of the overall sample.
This might be closer than it looks.
Doesn't SUSA have notoriously unreliable subsamples?
They do... or rather, they release lots of data on very small subsamples, so the odds are high that you'll get some weird data every now and then. The "tied among whites" part of the story has a higher likelihood of being true than the "Romney beats Santorum 6-1 among Oregon Hispanics," or "Hispanics are overrepresented in the OR Republican primary" bits.