The long-term vision - soft thinking or an essential building block?

I was thinking the other day about one of my last Board meetings of 2007. A year or so earlier, we'd agreed a broad-brush picture (aka vision) of what we thought the company should look like in 2020 - which will be long after the current leaders have all departed. This 'vision thing' is not uncommon in companies.
Then, at the end of last year, we found ourselves faced with a choice: either a strategy that would probably give an attractive return over the next four or five years, but which compromised the 2020 vision; or a direction that probably wouldn't pay for five or six years, but which potentially would contribute significantly towards that long-term vision.
It was a robust discussion; at the start I felt the majority mood was for the shorter-term option. In the end, though, and thanks to some outstanding chairmanship (that might be another story), the Board reached a unanimous position, sacrificing the short-term so that the 2020 vision remained intact.
To me, though, the real lesson from this was not whether we made the right decision - which I hope we did! - but how essential it was for us to have done that long-term thinking earlier.
If we hadn't done that, then I am almost certain we would have taken the superficially-
attractive shorter option. It's not that the long-term vision is necessarily right, but that, without it, we would have had no reference point to compare the short-term against longer-term decision. We could as easily have reached a conclusion for all the right reasons that the 2020 vision was flawed and have taken the shorter option.
As I said earlier, lots of boards develop long-term visions: I believe that one way to test whether it actually means something is how often you make hard choices, tested against that vision.
What do you think?