[zoobird]

I've read maybe a third of the book, and Bill (and I also assume he means Chen) does indeed spend plenty of time talking about how an unexploitable strategy is not always the most profitable strategy. However he also says that this is mostly true against bad players, and that against better players, the 'reads' that you're trying to exploit probably aren't as good as you think due to small sample size, players adjusting to conditions and learning, and a variety of other factors. Because of this, the book does encourage non-exploitable play as a general approach against decent players.