A nitpick and a rhetorical damper, in turn.
>The recent change from 17 to 20 years was
>to align U.S. patent law with GATT provisions. . . .
The U.S. patent term ran formerly from date of *grant* plus 17 years. It now runs from the date of *application* plus 20 years. That may or may not work out to be an increase in the overall length of protection, depending on how long one dithers in the Patent and Trademark Office application process. One nice thing about the change, though, was that it largely obviated the "submarine patent" problem, whereby a party would deliberately keep its application in the PTO's confidential halls so as to later spring it on an industry that had independently discovered the covered art.
>Without [the patent] mechanism, no research and development project
>would ever make it past the CFO of any company. . . .
Let's not exagerrate. A great deal of R&D aims at winning trade secrets, not patents. And a world entirely without patents would no doubt see the rise of alternative institutions, based in contract law, tort law, and corporate structures, to achieve many of the same incentive effects as patents. Then again, don't forget that R&D is not an unmitigated good. Perhaps patents *over* stimulate R&D.