I'm not sure why the "not with the Mac Pro" thing carries weight. Apple changed the name from "Power Mac" to "Mac Pro". Changing the name doesn't suddenly mean the entire product refresh rate horrifically shifts.
Except that in this case, the name-change did invoke a shift in refreshes. Here's the historical data:
Intel Mac Pros:
Today: 309 days & counting (10 months)
07/2010 511 days (17 months)
03/2009 420 days (14 months)
01/2008 279 days (9 months)
04/2007 240 days (8 months)
08/2006 292 days (9-10 months)
G5 PPC PowerMacs:
2005.10.19 (6 months)
2005.04.27 (6 months)
2004.10.19 (4 months)
2004.06.09 (10 months)
2003.08.18 (shipped) (2 months)
G4 PPC PowerMacs:
2003.06.23 (5 months)
2003.01.28 (5 months)
2002.08.13 (7 months)
2002.01.28 (6 months)
2001.07.18 (6 months)
2001.01.09 (6 months)
2000.07.19 (7 months)
1999.12.01 (2 months)
1999.10.13 (9 months)
G3 PPC PowerMac:
1999.01.05 (n/a)
Taking the averaging update info from the above:
G4: 9 updates in 53 months (jan99-Jun03) = ~5.9 months/update
G5: 5 updates in 27 months (Jun03-Oct05) = ~5.4 months/update
(FYI, averaging these two PPCs ... 14 in 80 months = ~5.7 months/update)
Intel: 6* in 67+ months (Oct05-today) = 11+ months/update
And the math indesputably shows, the Intel Mac Pro refresh rate has been "twice as long" as the history of the prior half decade of PPC PowerMac refresh rates before it.
But...the problem with this observation is that it isn't particularly relevant because today, Apple is using Intel CPUs and is thus reliant upon Intel's development cycles.
-hh