with a 1/4 pci divider... if they run their fsb at 150mhz... their pci bus is at 37.5mhz? if they run at 160, their pci bus is at 40mhz. what r the possible side effects (pro and con) of running the pci bus at 40mhz?
Yes those would be the PCI speeds and that puts extra strain on parts like network cards,hard drives, graphics cards etc but usually they tolerate it. I didn't worry too much about the PCI parts but when my hard drive failed catastrophically, i was kind of sad.

The funny thing was that I had only overclocked the FSB a few MHz at the time (133 to about 137) but the hard drive was an IBM 60gxp (sensitive to overclocking FSB). Who knows, maybe the overclock didn't cause the failure because the read head sounded like it hit the disk(s) and those 60gxps have a bad reputation. It's smart to back up your HD frequently.
The benefit of overclocking the FSB is of course the better data transfer rates and the fact that you can reach higher processor speeds than you can when only the multiplier is maxed out (only 12.5 on the 8k7a 12.5x133= only 1662). Also, if you used only the FSB to overclock, you might not have to unlock your processor if it's multiplier is locked (like the Athlon XP). I may have forgotten some benefits/risks but those are the main ones. Just overclocking by multiplier (no FSB) is a much safer way to go if you want to play it safe. Then just the processor will wear out a little faster but other parts aren't at risk. That's how i used to overclock and may do more of even if it means I can only reach XP 2000+ levels (on a XP 1600+ i'm getting to replace a 1.3 Athlon).