Consider: there is no room for improvement for the perfect scorers, and not much for those who missed just one or two questions. Won't the next history teacher who has them look bad if someone uses "value-added analysis" to measure her effectiveness?
In a way, it will probably be easier to show gains--value-added proof-- with students with Far Below, Below, or Basic scores the year before. Teachers of the gifted may not look so shiny with value-added scores. Do you think Usain Bolt's sprint times could improve as much as mine could?
Does the value-added analysis account for the fact that the higher the score the previous year, the less room for improvement there is? Dun dun dunnnn!