It couldn't have taken Apple engineers 2 1/2 years to rewrite the Finder, and I doubt they were sitting around doing absolutely nothing, so I wonder what the hell has been going on?
Most significant OS X version yet? Bleh. Going from 10.4 to 10.5 is more like going from 10.0 to 10.1 and it only took them 6 months to do the latter.
On the "surface" I have to agree. I like the "cleanliness" of the new look, but it's not a "2 1/2" year undertaking.
However, Leopard seems to be much less about what is on the surface than what is below the waterline. You can quickly point to the effort of doing a full 64-bit rewrite, but I think the new processing technologies are getting lost as we focus on the eye-candy. An apropos example in today's hardware market is Apple's push for multi-processing apps and APIs in Leopard. While the PPC boxes will likely enjoy the fruits of this effort, Intel boxes will likely see more, if there is any truth to the statements of how closely Intel and Apple are working together on making full use of Intel's technologies. Such Leopard advancements could yield significant boosts in performance, especially for those who actually make many by reducing production time (pros, not “hobbyists”).
In addition, there is another possible complicating factor to consider when critiquing the time-to-market of Leopard. I cannot help but believe that some/much of the Leopard delay is due to the switch to Intel (an excellent business move IMO). Sure, Tiger works well on Intel (Apple did an unprecedented job at making the transition VERY smooth), but I believe Leopard is Apple's first real Intel optimized OS. Owning both PPC and Intel Macs currently, and in light of many of the tests/reviews, there really isn't just a killer performance difference using the new Intel hardware, even with UB apps. Sure, some are better than others, but overall it's just like we switched tracks, not caught a newer/faster train. I guess the short version is the following analogy:
Tiger - “let’s make a great OSX for PPC that also works on Intel”
Leopard - “let’s make a great OSX for Intel that also works on PPC”
Certainly this is nothing more than my own conjecture, but it seems reasonable. So, with Leopard, I believe we will really start to see the Intel boxes come into their own, offering solid performance superiority to the loved, but outdated, PPC. For me, the wait doesn’t seem so long considering the above.
PS: I would have preferred to have heard greater detail on multi-processing, GPU optimization, and other performance improvements in the recent keynote vs. a rehashing of most of the same features from last year. Seems like the replay of last year's speech is drawing a lot of criticism