Haven't tried LR and the tests I did in CR where a solid version back so YMMV. However, when I compared the two, the differences in CR and LR where almost identical. I had to actually use calculations to see the differences at a pixel level. Thomas Knoll once reported the math IS different which makes sense. One is linear gamma encoded for one. My take is this. IF I know I need to rez up from the get-go, I'll do it in the converter. If not, I don't and will, if necessary do it later. But really, do a test. I'd use Bicubic smoother in PS. Its always more valuable to test the waters yourself and see if there's anything YOU think is useful in one technique versus the other.