The biggest gap between Nava and Sizemore right now has to be expected accuracy of projections. I'm not going to rehash the specifics that have already been laid out, but Nava is pretty understandable. He still has a range of possibility, but we should grasp his age, BABIP etc probably indicate any rational expectation should be below last year's performance, and his up and down in a reasonably narrow band.
What I don't fully get is the expectation that projecting Sizemore is as much of a science, and that the probability bell curve for him might have a peak lower than Nava, but it is considerably wider and hazier.
I don't think it is unreasonable or very unlikely that he would not be able to shake off the time off and really be healthy enough to be an above average major leaguer. But it is really difficult to make a case for this based on statistical projections of either his 80 PAs this year or his 435 PAs in the last four seasons with two injury plagued seasons followed by two completely inactive years. I don't think there is a reasonable basis to trust a statistical analysis and model of why he will be bad.
I also don' think it is unreasonable or very unlikely that he might actually be all the way healthy, and shaking off the rust and timing to get back to what you would expect from the 31 year old season of the player he was from 22 to 26 before injuries derailed him. I'm not throwing Roy Hobbs at him, but sometimes a career gets off track, and if health and age permit, there isn't a fundamental reason it can't get back on track.
So with two pretty substantially ranging realistic possibilities, I don't think Sizemore is accurately projectable without continuing to watch him play, and see if he is actually healthy or not (and how close he might be in health to the hypothetical 31 year old who didn't lose four seasons), and if he is able to overcome such a long time away.
With Sizemore having such a difficult to predict and wide range of possibilities, it is really difficult to gauge which side of Nava's likely performance he falls on without either waiting and seeing as statistics accumulate, or relying on the inexact old school eyeball scouting. I happen to love Nava for the narrative, but think Grady has a greater than 50% chance of outperforming him at the plate in 2014, and even more significant chance of outperforming him in 2015.
If the pool of 2014 data continues to grow without Sizemore showing more hints of being an 800OPS player, the Red Sox might force their hand at some point and give his playing time back to Nava. The upside of Grady being good Grady is significant enough to ride out a larger sample over a slumping and likely BABIP and age adjusting projectable Nava, which I think is exactly what is happening and should be expected to continue as long as he is showing signs of progress at the plate.
But there just aren't enough comparable career path's to really figure out what to expect from 2014 Sizemore based on 2005 to 2009, or 2010 to 2013 or 2005 to 2013. Mysteries that statistical analysis can't pierce are pretty unpopular around here, and our track record of failed reclamation projects probably tempers the optimism further.