I wonder could that be the real speed? With manufacturing tolerances and other factors. I mean why print 55.80mhz on an oscillator or processor ..just round down to the nearest and print( I mean on the actual chip).. There is always a + or - % either way when it comes to electronics. Could be SysInfo is pretty accurate...with somethings.
I'm probably way off so forgive me it is early. 
Rich
There are lots of factors on oscillators.. such as ripple,clock jitter and other things. tolerance is important too.there is also trace lengths on boards to consider and interference from other parts near by. frequencies that are not rounded off are used to fine tune in some cases.
When processors are made they dont know the actual speed they will run at until they got through testing. the parts will usually pass different speeds reliably so they get marked what speeds they can run. motorola for example mark chips with non working MMU/FPU as EC,and parts with fully working mmu/fpu are XC or RC. same thing with speeds.
There are many needs in the world for oscillators that are not "rounded off"
like radio frequencies for example.. just think of radio stations for example.
106.9,98.3 etc..dont forget ntsc also at 15.734khz
more than likely you may have a 55.80mhz oscillator in the warp engine.there is many factors in a computer that can determine what the chip really see's for a clock.Don't assume sysinfo reads it correctly in all situations
By changing the oscillator to something faster than the chip is rated you are effectively overclocking it. You could possibly get a little more speed out of that 25mhz 040(which the warp engine guys already clocked to 28mhz,and apparently someone tried to clock it a bit faster,it may not of wanted to go 30mhz so they backed off to the odd value oscillator you have to get it to run reliably. in theory you could put a 40mhz 040 in it,and stick a 80mhz oscillator in and have a 40mhz warpengine(keep in mind they are dividing the clock by 2 here and that is what the 040 is getting).
Mech
