<< why spend 255 on xp2000+
when you can spend about that much on P4 1.6a + Mobo for faster setup?
even with moderately o/ced spec of 1.6a... you can compete with xp2000+ >>
I have to agree with that, I don't understand why people are willing to spend roughly $100 more for the 2000+ over say an 1800+ when the clock difference is 1666MHz vs 1533MHz, a whopping 133 some odd MHz.  Paying $100 for the extra Hz will get you the extra performance, but for $100 you have almost enough to buy another 1600+-1800+ CPU.
But spend on an Intel setup?  Sure you could get great performance if you do over-clock it.  Some people aren't willing to nor are some people even educated enough to pull off such an over-clock.  Also, I don't see much hope for intel after the 1.6A and 1.8A.  The 2.0A and the 2.2s definitely do not come close to having the same over-clocking potential as the 1.6A or the 1.8A, so how will one upgrade for the future?  Well I don't see a solution.  The large over-clocks are achieved by stepping up the fsb speeds from 100 which allows for big results in the end.  But intel will be switching the P4 over to a 133x4 fsb.  This means newer processors won't gain as much of a performance increase by over-clocking the fsb from 133.
On the other hand, AMD consumers have the T-breds to look for when the time comes to upgrade.  A .13 micron fabrication process did wonders for intel and their Northwood P4s, I truly expect to see similar things happen to AMD when the Athlon gets a .13 micron core.
So IMO, AMD still stands as the better overall deal.  I was tempted to look into getting myself an intel Northwood setup, but one either needs to spend big $ to get performance from intel or take some risks.