This isn't going to be the most profound or penetrating insight I've ever served up (primarily because my baseball neurons and synapses aren't really firing at full capacity just yet), but I did want to take a moment to address something that, in the context of the C.J. Wilson-as-a-starter debate, strikes me as a misconception of sorts: the idea that Wilson, solely by virtue of his career-long inefficiency in terms of pitches thrown per inning as a reliever, should have his rotation candidacy terminated.
Dating back to the 2006 regular season (which, conveniently enough, was the first season in which he pitched exclusively out of the bullpen), Wilson has averaged 17.7 pitches per inning; this is a fair-sized tick higher than the average mark posted by his qualifying bullpen-exclusive peers over this same period of time, which is right around 16.5 pitches per inning. There's no refuting that Wilson has been below average in this regard, but a big problem arises when one goes so far as to assume that Wilson, if successful in locking down a rotation spot, would continue to average 17.5-plus pitches per inning, and thus would not be viable as a starting pitcher over the long haul.
Where this assumption quickly breaks down is in the reliever-to-starter translation; the key point is that relievers inherently use more pitches per inning than starters (by a factor of 4-5 percent in recent seasons, in fact), a phenomenon that can largely be traced to relievers walking and striking out more batters than their starting counterparts. What this means, in all likelihood, is that after properly regressing Wilson's peripherals in the bullpen-to-rotation transition and accounting for the adjustments in his approach, he'd likely use something more along the lines of 16-17 pitches per inning as a starter -- not necessarily average, but probably viable enough. It wouldn't concern me too much.
Now, as David Brown pointed out back on November 16th, the principal question we need to be asking ourselves is whether the value added in moving Wilson to the rotation would be greater than the value lost in downgrading from Wilson to the team's eighth-best reliever; right now, that's probably somebody in the vein of Doug Mathis (strong 2009 numbers, but lacking the repertoire and ability to miss bats required to sustain them), Pedro Strop, Warner Madrigal or Ben Snyder, the latter three of whom are young and loaded on talent, but are either untested or very light on major league results.
The problem with Wilson isn't necessarily one of lackluster projection as a starting pitcher. Quite the contrary, in fact. On a per-inning basis, I can readily see the No. 2 starter upside (player labels be damned), and I would imagine that management can see it too; otherwise, acquiescing to his wishes for a shot at the rotation is a waste of everybody's time, unless it was a move made with the sole intention of placating him. No, the problem is that he might only be good for 120-130 innings, at most, and there's enough risk involved in the bullpen-to-rotation transition itself that you find yourself wondering if it would really be worth all of the hassle in the first place.