why is the norm for windows systems about 2.4 Mhz to 3.0 Mhz or higher?Compared to unix systems it seems that unix does more and more stabel with less Mhz?why??!

I hope I'm understanding your question right.
Also, as it is a bit more difficult to give a "generalized" answer to this, I am going to assume that you have background knowledge in PC CPU history -- if that's not the case and if I'm not clear, please tell me so:

I understand you are asking about the "Megahertz gap" between the PowerPC family of CPUs and current x86 compatible CPU families.

First, this doesn't have anything to do with UNIX; it doesn't have anything to do with the question which OS you are using. The "Megahertz gap" is about differences in CPU architecture, and that's a hardware issue.

If you remember the 386/33 MHz and 486/33 MHz, nobody would have assumed these two processors to be equally powerful, even though both processors did indeed run at the same clockspeed (33 MHz).
This is just a very clear example showing that MHz is only ONE indicator of CPU speed.
It is true that a 386 at 33MHz was more powerful than a 386 at 25 MHz.
However, a 386/33MHz was LESS powerful than a 486/25MHz, even though it was running at a higher clockspeed.

To know the power/speed of a CPU, you have (at least) to know both the MHz AND what kind of processor it is. MHz alone does only tell you half the story (and is potentially very misleading). There are many more factors that affect CPU performance, but most would argue that for raw CPU speed, the CPU type and its clockspeed will be the most important info.
Note that this even true WITHIN the same architecture (386 and 486s belong to the same architecture, that of x86 compatible CPUs).
Between different architectures, the differences can be even greater.
A 68k architecture processor at 40MHz would perform very different from a x86 architecture CPU and a PowerPC architecture CPU would perform different still.

The problem is that very recently, there was a time when most recent Intel CPUs performed similar enough at any given MHz, so it was favourable for Intel to tout MHz as the "major" performance indicator. This led many to wrongly believe that MHz was the sole performance indicator, which it isn't.
It also drove Intel's competition crazy as it made e.g. PowerPC CPUs (which are generally, for reasons of hardware architecture, more powerful at a given clockspeed) look much more behind than they in fact were (for a while).

This misunderstanding about the relevance of MHz is however becoming unfavourable even for Intel now, so they're moving away from that, see http://arstechnica.com/news/posts/1079725353.html .

Again, though it can be difficult to make cross-architecture comparisons, PowerPC CPUs are currently faster at a given clockspeed than x86 CPUs. So if you had a x86 compatible CPU and a PowerPC CPU of about equal speed, the clockspeed (in MHz) of the x86 type CPU would be higher.

Coming back to UNIX:
Compared with the Windows family of Operating Systems, most UNIXarian Operating Systems make more efficient use of hardware resources, so the overall system speed would indeed be quicker when using a UNIX type OS, and that applies to whatever CPU you are running the OS on. (Windows nowadays mostly only runs on x86. RISC Alpha CPU support used to be a big thing, but it's pretty much gone the way of the dodo.) UNIX on x86 will likely be quicker than Windows on x86.
NB: You can run UNIX style Operating Systems on pretty much all CPU architectures in existance, see http://www.netbsd.org/ .