To be honest, NBC & Variety didn't convince me. I already felt that way.
In a vacuum, I agree with Bill down the line that absolute ratings matter, and winning and losing are just for chest-thumping PR purposes.
Even outside the safety of the vacuum when the margins aren't tiny it's still easy to call. NBC would rather Leno lose with a 2.0 adults 18-49 than win with a 1.0. You don't need to think twice about it. But then the NBC sound bites would be "there's been dramatic growth in late-night viewing and it's been a boon for everybody!"
Sadly, and not just for NBC, that isn't the case.
Would NBC rather win with 1.0 for Leno vs. 0.9 for Letterman than lose 1.2 vs. 1.1? Bill suggests NBC would absolutely prefer the 1.1 to the 1.0, even if it meant Leno was in second place. But at those tiny margins I not only think it's possible that Bill is wrong, I think it's quite likely.
Bill fairly notes the egos involved. This is a case, especially given those egos, and NBC's recent history where the PR aspects of winning matter more than normal. At the tiny margins were talking about the absolute ratings, and their ups and downs, might not matter as much as Bill suggests (edit: the year over year drops aren't so tiny and definitely catch the attention of, and cause some agita for the NBC brass).
Bill also suggests NBC ad sales don't sell wins, but rating points. To clarify, I think what he meant to say is NBC isn't paid on wins, it's paid on ratings points. There's no doubt that for the NBC sales team "Leno winning" is a selling point.
But no matter how you slice the whole Leno out of late night to primetime back to late night experience, it has been a disaster for NBC. But it would've been perceived as even more disastrous if Leno wasn't at least winning. Out of a simple desire to avoid that perception and the accompanying headlines, a win with a 1.0, looks better than a loss with a 1.1.
Bill thinks there's no reasonable argument to be made that winning matters. We disagree.