I'm not going to say that benchmarking is going to make a difference or anything, I'm just inclined to use them because they give a good indicator as to how well your rom/device is using the data/drivers/settings/etc installed on your device. I look at a benchmark score and see that "x Rom" is performing at a better rate than "y Rom". I might like "y Rom"s functionality and usefulness more, but if I can see that "x Rom" functions better, I can try to incorporate the benefits of X into Y. The usefulness of benchmarking is to see what my device is making use of the information/software/etc of the ROM it's on, so... IMHO, benchmarking serves a great purpose... Just gotta know what to do with the information it provides.
The reasoning as to why I posted the scores that I did is because the "a-typical grouper rom" posts scores in maybe the high 5k to low 6-k area with my normal set of settings (without ART enabled), and usually in the high 6k to low 7k area with ART on. With this rom, I posted 8.4k, so I was impressed and wanted the community to see how well this ROM functions.
Cheers.