I think you definitely want 1-3 mid-pricers to allow for the swaps (currently have Williams, Lynch and Fisher - although Williams is closer to a top rookie than true mid-pricer).This can only work if the Rookies are scoring well and not under 50 and also it's best 18 so MPM might be an advantage. Also your theory of trading in Midpricers after R2 means either trading out premos to get them or starting with lots of cash in the bank.
Starting with Midpricers if they are a bust are an easy downgrade to a rookie who might have been missed or possibly trading a couple 1 up and 1 down.
Lets say you go with a 12 premo, those 3 mid-pricers and 15 rookies (structure allows for all the top ones). In the rounds that Williams and Lynch play, if you have one premo from each of the early byes, you will line up with 11 premos, 3 mid-pricers and be able to loop at least one rookie on to have 9 rookies total.
This means you get the best 4 rookie scores out of 9. It is very unlikely that all nine of those rookies score less than 50. Looking at my team last year and the rookies on field (plus two bench due to one less premo and one loop):
Round Two: 114 (Sheezel), 105 (Ashcroft), 78 (McKenna), 76 (Ginbey), 73 (Caminiti) 71 (Chandler), 68 (Wilmot), 67 (Constable), 55 (Goater), 52 (Mackenzie) = Average of 93.25 for the on-field rookie scores that count towards your round score.
Round Three: 120 (Sheezel), 104 (Ginbey), 95 (Baker), 90 (Chandler), 89 (Mackenzie), 80 (Ashcroft), 69 (Wilmot), 57 (McKenna), 46 (Goater) = Average of 102.25 for the on-field rookie scores that count towards your round score.
Round Four: 111 (Sheezel), 86 (McKenna), 75 (Ashcroft), 73 (Greene), 67 (Phillips), 59 (Chandler), 50 (Wilmot), 49 (Ginbey), 40 (Mackenzie) = Average of 86 for the on-field rookie scores that count towards your round score.
While these are only three rounds, and each year will differentiate, think it is a decent example of how high the rookie scores can be for the top four rookies each week.
Now what MPM does it usually take away one premo and one rookie and replace them with two mid-pricers (roughly). This means you lose the stability of a premium player and add in two mid-pricers and also get to replace the worst rookie. However, unlike previous years, you aren't dropping out a 52, 46 or 40 (the worst rookie scores in each round above) you are dropping out a 76, 90 and 73 and so the points gained aren't as high.
So, in my mind at least, you are taking on the extra risk and volatility of mid-pricers, but you aren't getting as much of the benefit of the additional scoring in those four early bye rounds.