Thanks everyone. I was expecting massive disagreements, lots of shouting and insults. Pleasant surprise.
Davos, I would argue that you can approach performance tuning scientifically. No, it doesn't boil down to a single variable and no, production is never going to be the same as dev. You can still, in your testing environment, fix as many of the variables as possible and run the tests, make the comparisons and draw conclusions.
It's pretty unusual that a slower query use less resources. Unless it spent that extra time waiting for something, that extra time was time running, using resources. Plus a slower query will be holding locks longer, possibly interfering with other queries for longer, etc.
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
We walk in the dark places no others will enter
We stand on the bridge and no one may pass