The book has changed over the years and the 2012 edition is different than previous years. The player profiles are larger (there's 7 players/page rather than 9) and there are more statistics and skills ratios. The Encyclopedia of Fanalytics now takes up the first third of the book. It's been improved and is a fun read on its own.
The authors discussed the accuracy of predictions quite a bit over the last couple years and now they have fully embraced a weighted 3-year average unless skewed by obvious trends, scouting reports, injuries, and/or age. Their point, and it's valid, is that the inherent error in actual performance is so great that accurate predictions are impossible. Think about accuracy vs. precision and standard error, or noise, of the data.
Twenty-five years ago when this book was first published, there were very few predictions available. The internet didnt exist. When we drafted our Roto teams, we used the baseball annuals published by the Sporting News (I miss the Baseball Guide and Baseball Register), insight from Bill James, and the spring training magazines. By the late '80's pioneering player valuation authors like Shandler, Patton, and Mann-Manley slowly became available at bookstores. Now, anybody can get pretty good projections the night before the draft.
Shandler's point in this book, however, is that such precise projections are not really that accurate. The Forecaster gives you the tools to do your own evaluations. Use them to select the players you think will perform the best. In this sense, fantasy baseball has come full circle if you're doing your own work. The way to beat the parity created by a plethora of available information is to gather tools that provide the edge of accurate insight.
And finally, are the 60-word player write-ups less pithy than the old 30-word gems? The Matt Latos bio includes a description of his career trajectory, "He came quickly but proved to have staying power". No further comments...........