Exactly! Plus without printing a small sample of the two options and viewing them side by side, forget about 'evaluations' of such applications. I've done some of the work recently (Photoshop versus Lightroom versus Perfect Resize for a Webinar) and on an 8x10 print of a sample of the full image, PR didn't bring anything useful to the party, took a lot more time and after a demo times out, money. With proper capture sharpening, LR did the best job with Photoshop pretty close behind.
If your judgement was based on PR resize vs Bicubic smoother, and the magnification was relatively small, I can understand. However, the OP is looking for some serious upsampling amounts
. He also expressed a favorable opinion on Blow-up, which means he has used and compared that to form an opinion for the intended goal, which is huge
image sizes. Genuine Fractals, or Perfect Resize as they call it nowadays, used to advertise 4x enlargements as giving good results. My opinion, based on print comparisons, was that it tended to posterize the image data too much, thereby losing the smooth organic gradients (e.g. OOF blur), compared to e.g. Photozoom Pro.
An up to date comparison between the two (and Photozoom Pro) could be interesting, but Perfect Resize crashes on my computer so somebody else will have to give it a try. I could do the Photozoom Pro resample if needed, it's running just fine, and can do much larger enlargements (e.g. 10x or more, it's limited at 1 million by 1 million pixels).
ANY review of these kinds of products that try to convince you which is better, based on what they show on screen is a review to ignore. You have to print the various samples at size and look at them.
While I agree that an actual print is the best comparison, anybody with a bit of experience can judge the expected output difference based on a crop on display just fine.