my limited understanding is that all mice have prediction and smoothing. some mice have more then others, and its typically excessive prediction and smoothing that is undesirable. Smoothing is typically a byproduct of trying to counteract the jitter created at high dpi, but it is not as simple as high dpi equals high smoothing. The 2013 deathadder for example its reputed as having some of the heaviest smoothing and has a max dpi of 6400, while the mionix naos and avior 7000 has 7000 dpi and less smoothing. Mice with the 3090 sensor, that have over 3500 dpi are also reputed to have very heavy smoothing.
Logitech claims that with the sensor in the g502, a responsive and accurate sensor were their primary goals and they only had the max dpi at a range that didn't compromise either of those goals. Of coarse if companies were always 100% honest the world would be a completely different place. Also, my understanding was that they claim all their mice with delta zero technology have that same standard, but their g400s has delta zero technology and is reputed to have fairly heavy smoothing. And in my experience in the world of peripherals, engineers often underestimate what the human body is capable of. Engineers might think a couple ms is negligible and people might be able to notice it. Also with mice, it seems like marketing departments are pretty convinced that higher DPI sells more mice, so its likely many engineers are pushed towards exceedingly high DPI despite it having a negative effect on the mouse's performance.
One of the biggest problems with smoothing is that there is no clear test for it at least there is none readily available to consumers. You can't draw a line or turn in a shooter and through a video of that action people can see smoothing. Because there is no clear test, its all feel, and when its only about feel, the placebo effect is extremely real and extremely significant. Roach for example is one of the people that is most sensitive to smoothing, but through his experiences he is also convinced that High DPI results in smoothing, so while he may legitimately detect smoothing in many mice, there is the threat that he could test for smoothing in a high dpi mouse and because he is already fully expecting there to be smoothing, he detects smoothing that doesn't exist.
Because testing for smoothing is unreliable because of the placebo effect, we have to rely on numbers and because the g502 is brand new, it is hard to conclude whether it likely has significant smoothing or not until more sensitive people test it for themselves.
Edited by Atavax - 4/14/14 at 6:38am