Soltis examines much of the conventional wisdom regarding the value of the pieces, subjecting it to critical scrutiny. The book is well researched and grounded in the history of chess analysis. Chapter one, for example, runs through several schemes for assessing the value of pieces (such as the common knight and bishop = three pawns, rook = five pawns, queen = nine pawns), dating back to Shatranj (when the piece that would eventually become the queen was much weaker) up to efforts of chess software programmers struggling to create engines that understand the value of the bishop pair.
More than anything, the book offers lots of thoughtful reflections upon and examples of the imbalances that are the heart of strategy and tactics in chess.
"Piece value begins with mobility." This sentence begins chapter two, which then looks at cases when a doubled pawn loses its value, then moves to good and bad bishops and knights, and on from there.
Chapter three looks at board range (the size of the board) as it relates to the value of the pieces. We all know, or should, that bishops are better than knights when there are pawns on both sides of the board, and that a knight is often equal to a rook when the pawns are all on one side. Soltis takes these observations deeper.
Rethinking the Chess Pieces is, as the name implies, a book for those looking for instructional material for reflection. It's not a book for teaching tactics like pins and forks, but for understanding why and when a given pin or fork might lead to exchanges that confer a strategic advantage for one player or the other.