My problem with all of it is that nobody seems to want to stand behind their work...
You have Nate Cohn at the NYT commissioning polls showing a relative neutral environment (House polls in KS, NM, NV, Senate polls in PA, NV, AZ, WI) and the writeups are about how they may all be wrong.
You have RCP building a model which deducts share based on previous polling misses in the last three cycles, yet still is willing to project a different result than what the model shows (see MI Gov).
And then you have Nate Silver, who built a website based on modeling this stuff yet basically seems more interested in punditry and, similar to Cohn, looks for reasons to doubt his own work.
None of this matters obviously, Tuesday will come and go and very little that these guys do or don't do will change any of it. But the lack of spine to stand behind their work is kind of pathetic IMO.