There are not enough average pitchers to cover 5 rotation spots. In fact, there are not enough pitchers that pitch 100 innings to cover 5 rotation spots. By the time you get to the 5th spot in the rotation, you're really talking about a replacement pitcher, perhaps shuffling some minor leaguers, a midseason acquisition, or perhaps an injured starter that gets it together at the end of the season, etc. 

If you think about it, there is only an artificial draw for a 5 man rotation, in terms of talent. The number of innings a team needs to fill to complete a season is pretty much set, with some wiggle room depending upon whether or not a particular team is winning more regulation games at home, or whether another team has more extra innings games, etc. The necessary workload for a team does not change depending upon the size of a rotation; that merely adds another arm to the equation. The result is that despite developments in sports medicine, and new interpretations about how an arm should be resting, the invention of the 5th starter did not fill a natural need in the game of baseball. In fact, if anything, the invention of the 5-man rotation created a false scarcity of pitching (along with the left-handed relief specialist, which further divides the team's IP workload into specific tasks that does not do anything except for create the need for more pitchers).

Subsequently, there is no natural progression through the rotation, based on performance and value. 

If you look through 2004-present, using the parameters that I did in the previous post, you'll find that -- based on value -- there are not more than 10 #1 pitchers in the MLB; if, of course, you take #1 pitchers to be the best pitchers in the league, the most elite; something like 60.0 VORP and higher, a type of pitching season that not many accomplish (after the 2009 season, there will have been something along the lines of 320 60.0 VORP seasons from 1954-present, basically around five or six 60.0 VORP pitchers per season). 

Similarily, there is an over-abundance of low-rotation pitchers. If you follow the VORP-production chain incrementally, from 60 to 0, you might think that #1 = 60+, #2 = 45+, #3 = 30+, #4 = 15+, and #5 = 0+. The problem, of course, is that on this scale there are actually six rotation spots, when you add in the number of pitchers that are below average, or -0 VORP. In this case, there have been many more sixth starters than first starters in the past six seasons. Usually, twice as many pitchers post below average seasons than 60.0+ VORP seasons.

This skews the rotation map, which cannot be drawn according to actual production and cover all 30 teams (using only +100 IP pitchers):

2009: 9 one, 13 two, 21 three, 33 four, 22 five, 23 below average

2008: 5 one, 16 two, 34 three, 34 four, 25 five, 26 below average

2007: 7 one, 21 two, 25 three, 33 four, 37 five, 16 below average

2006: 7 one, 15 two, 28 three, 38 four, 28 five, 9 below average

2005: 8 one, 12 two, 35 three, 28 four, 35 five, 22 below average

2004:  9 one, 14 two, 23 three, 46 four, 38 five, 14 below average

As you can see, there are not nearly enough #1 and #2 pitchers to cover 60 rotation spots. Not close. This is why it ismisleading to judge teams according to their own rotation spots; not every team has a #1, #2, #3, #4, #5 starter distribution, and subsequently, a team's #1 is not necessarily a #1 overall. 

What I find astonishing about this phenomena -- that MLB teams use a pitching rotation strategy that creates a false scarcity and does not come close to providing them even service from their pool of pitchers -- is that it's not simply an outcome of expansion, or the invention of the 5-man rotation itself; the number of 100+ IP pitchers per season has remained consistent since the 1960s, and even the number of 200+ IP seasons has remained rather consistent:

1999: 148 100+ IP

1989: 141 100+ IP

1979: 148 100+ IP

1969: 139 100+ IP

1959: 87 100+ IP

Even in the pre-expansion era, the same basic percentage of pitchers pitched 100 innings in a season; this is a testament to the fixed workload of MLB teams, throughout the decades. No matter how you slice the rotation, there's only a particular number of work that a pitching staff needs to finish.

The same goes for 200+ IP seasons:

1999: 41 200+ IP

1989: 48 200+ IP

1979: 54 200+ IP

1969: 58 200+ IP

1959: 37 200+ IP

I am by no means suggesting that losing nearly 20 200+ pitchers over the span of four decades is insignificant; but, given the development of the 5-man rotation, you can tell that the consistent number of 100+ IP seasons simply reveals a redistribution of IP, be it through an insanely overworked relief ace (in the late 1970s, it wouldn't be strange to find 3-5 relievers in the Top 30 VORP seasons among pitchers with 100+ IP), or through low rotation starters in the 2000s. This reconstitution of pitching roles explains some of the shift away from 200+ IP while maintaining a steady number of 100+ IP seasons.

The development of the 5th rotation spot is not natural, does not cover a need, and therefore no division or measurement of 5-man rotations will be satisfactory. Even if we reconsider how to judge the production of 5-man rotations simply according to team needs, rather than actual performance, we have a rather unsatisfactory result. Dividing the 100+ IP pool in 2009 by 30, for instance, we find the following production range:

#1: 39.0 VORP+

#2: 22.0 VORP+

#3: 5.0 VORP+

#4: -22.0 VORP+

And of course, there are not enough 100+ IP pitchers to cover a fifth spot for the league -- not even close.

Compare this with our earlier chart; if we need 30 #1 starters, the VORP range drops to 39.0-78.0+; naturally, though, there are only 22 pitchers that even reached the 45.0 VORP range in this league. This means, of course, that there is actually a terribly wide range in top rotation production, and the same can be said of low-rotation production, which suddenly includes pitchers that are slightly above average, as well as the worst pitchers in the league.

According to actual need, there is no good way to judge rotation production if we actually want to divide it into spots. The conclusion, of course, is that runs allowed is all that matters, and how the workload is divided should be dictated by the type of arms available. I have said it before and will say it again; some gutsy team that picks up veterans in the mold of Javier Vazquez, Mark Buehrle, Carlos Zambrano, Dave Bush, Jeff Suppan, etc., and limits their rotation to four spots (with a fifth starter only used in cases of necessity, rather than as a regular spot) will raise some eyebrows but probably produce a rather strong rotation. All it will take is one team to successfully move away from the 5-man rotation for a couple of seasons to show that it is not necessary, and creates a false scarcity of pitchers, which will become evident when teams get tired of continually seeing Jeff Suppan, on the one hand,  match up against Tom Gorzelanny and Kevin Correia on the other.

Similarly, it is unrealistic to expect your MLB team to come up with #1 and #2 level pitchers in the off-season; in the wide range of pitching types created by the 5th starter system, having good #3 pitchers that can exceed their previous production every now and then is just as good as having a #1 or #2, especially given the rate at which true #1 production is actually repeated by particular pitchers (recall that of the 96 top pitching seasons from 2004-2009, only 16 pitchers have actually repeated their top level of production for two consecutive seasons). 

Acquiring #1 and #2 pitchers will not become a problem until teams regularly return to the 4-man rotation; and in this sense, I argue that it's time for a team to acquire veterans that can handle a workload, and maybe a developing young arm or two (who will be regulated more than the veterans), and pitch a 4-man+swingman rotation style. 

The point is that there is more than one way to build a rotation besides going after #1 and #2 pitchers, and one of those ways is to cut oneself off from the issue of false scarcity itself. One of the best ways to subvert the five man rotations is not to go out and acquire top talent, but to align your team according to a completely different pitching rotation.