If the output distribution follows a bell curve, which it likely does, there are probably already a significant number of R bins in existence, just not enough to offer them in volume. It'll probably take a little longer to tweak production so the bell curve peak shifts to a higher bin. It's not like Cree purposely tries to hold back or release bins to follow a set time line. Rather, as the process is tweaked more and more LEDs fall into higher bins. Right now I'd say Q2 through Q4 is where most of the production LEDs fall, with Q3 probably being the peak of the bell curve. P4s and Q5s fall on the outliers, meaning that the yield is less for these bins. Q5s obviously command more money than P4s even though both are likely made in similar numbers at this point. Further out, you would have P3s and R2s, perhaps even a small number of P2s and R3s. None of these are made in large enough numbers at this time to offer except possibly as engineering samples. No way of knowing exactly when the R bins will become common enough to offer, but I'd say sometime before 2008 based on previous trends. It seems production is advancing by maybe 3 to 4 bins a year. This time next year I'd say R2s or R3s will be the most common bins, with R4s, perhaps R5s, available in smaller quantities.