I'd suggest that a number two starter pitches fewer innings than a number one starter and more than a number three starter.
Using that definition, I quickly checked (source THT) the 15th through 28th-ranked AL pitchers in IP for 2007. Here are the median stats:
W-L: 14.5 - 10
ERA: 4.24
SO: 124
BB: 63
IP: 197.5
GS: 31
OK...a number two starter might be a guy with a 14-10 record in 30 starts, an ERA a bit over 4.00, a 2:1 K/BB ratio in MLB, who approaches 200 IP.
Ian Kennedy isn't near that level today. I cited one report suggesting that he's near his ceiling--I've read others. I also know from reading Pitch f/x logs that he struggles unless he's given the outside pitch, especially with respect to lefties.
Some might consider the nine-game 2008 sample too small to assess that Ian Kennedy won't reach this level. I don't. My reasons are these:
1) His velocity is low for a young pitcher, maxing at about 92 mph. Such pitchers can succeed, but there are few who do. Greg Maddux was the primary exception, and it's now considered that his being given an expanded strike zone by most MLB umpires was critical to his success. With Pitch f/x in every ballpark, preferential strike zones are becoming things of the past.
2) If one checks the xFIP of Ian Kennedy's work in 2007, it wasn't much different from his xFIP in 2008.
[table] Season | ERA | xFIP
2007 | 1.89 | 5.15
2008 | 7.41 | 5.86 [/table]
The "real" Ian Kennedy is probably around that 5.86 xFIP, which is what he's getting with a fair strike zone in most of his games. That would leave his ERA in the high fives, much worse than the norm for AL second starters. If Kennedy had room to grow, things would be different, but pitchers as a rule peak earlier than position players, and Kennedy in particular appears to have hit his ceiling.