I have seen a marked improvement in overall scrolling, swyping/SlideIting, slideshows etc. While it's hard for me to see a remarkable, discernible difference in video playback (not saying the FPS aren't increased) but I notice more that the overall performance of graphics related tasks are much smoother. That's where I see an improvement.
You're not going to see any difference at all, in fact, in raw video playback. That's where HTC tried to make their defence about this whole debacle before hand. Yes, all VIDEO playback is 24-30 fps. Ask anybody who plays their games on a PC, there are MAJOR differences from 30 - 60 fps. I can vouch for that. Back onto the subject though. Video playback may not be affected by the FPS boost, sure, I'll give HTC that much. However, everything else, and I quite literally do mean everything, is affected by being capped at 30 fps. It looks choppy, it stutters, and it makes you want to put the phone down right where you found it. They can sit there all day claiming it's perfect because videos playback at 24-30 fps, but I don't turn my phone on to watch videos until the battery dies. I generally intend to use it. If it's laggy, can't keep up with me, and stumbles through every task I throw at it, how exactly is that perfect?
To those people not being able to see ANY difference after the update, tell me, what phone did you have before this? If you had a flip phone, or some cheap free-with-contract texter, I can give you credibility, because you don't know what to look for. Anything you did on the Evo looked SUPER smooth compared to before. Go pick up an iPhone or even a palm pre. It can, and will put the Evo to total shame in graphical performance. I know, i know, go ahead and state how much faster the Evo actually is... i know this. But, before the update, it looked like crap while it was going so fast. Imagine a drunk race car driver sitting in a formula one car racing a sober driver in a BMW standard luxury car. That's essentially how it panned out in my eyes, before the update/rooting.