It's as though you're trying to pick apart my position on this, while I just don't feel a need to warp reality to make you happy. But maybe I'm being too blunt for you with my "rule of thumb" recommendations. Try this on:
Running unnecessary things on the central database of your computer can increase the risk of problems. Whether the added risk of doing what you like instead of what's needed could be considered substantial is up to interpretation. We're all free to do what we like, but I stand by my statement:
You simply don't need to run 3rd party software on your registry.
Here's a year's worth of overall system benchmarks on my main Windows 8.1 workstation that's had daily hard use (24/7), and without my using any magic registry performance enhancers.
Do you see any negative performance trend here? I don't.
As the user of this particular system, I am not finding it to be any less responsive or powerful - with one small exception - than it was the day I installed Windows.
The small exception I'm compelled to describe for the sake of completeness is that Microsoft specifically wrecked performance of the Direct 2D subsystem with a Windows update in November 2014. They briefly un-wrecked it in February 2015, then wrecked it again with the very next updates. Note the radical changes in Direct 2D performance in November 2014 and February 2015:
I am not alone in seeing the above.
The above makes comparison of overall benchmarks since November 2014 difficult with those from before, because the one Direct 2D reading affects the overall score a fair bit, but here's the trend after that time.
Factoring out the Direct 2D anomaly, which as far as I can tell doesn't actually affect anything I personally run other than the benchmark program, you can see my performance has not been degrading in the recent 7 months as well.
I also have archived results showing a similar lack of degradation with Windows 7 systems if you're interested in seeing them.
Let's have this discussion move to more objective measurements and results, shall we, and stop with the nitpicking of terminology. I'm certainly willing to be educated by other peoples' actual results.
Edited by NoelC, Today, 12:28 PM.