I'm not engineer, so I can't say with 100% certainty, what I can say though is I wouldn't think checking for errors is a waste of resources personally. Another thing I can tell you when I'm overclocking and use IBT to stress test, I know exactly when I'm on the verge of failure and that's when my next OC bump yields in no gains or even a slight loss in gflops, even if it makes it through the test. Sure enough, another MHz or two on the base clock throws an error on the subsequent run.
First we need to establish that your stuttering is due to your overclock. Do you only get stuttering when attempting to overclock or does it happen when everything is at default settings too? I know you said you got 60 before and 45-55 now but that's not at all a scientific measure if all you did was go into a random game and looked at your fraps counter. This is one area where a static benchmark is helpful. If you have a game with a built-in benchmark mode, use that. If you don't, use something like heaven, but lower the graphics setting to minimum, or better yet, cinebench. These will be repeatable tests and if you are in fact losing performance when overclocking, it will show up in these tests.