HOT lanes are simple in nature compared to tolling an entire portion of the city. Volume at entrance and exit is one dataset, what about everything in between? The grid is massive, is just checking toll gantries sufficient? Some may say no, I'd tend to agree. Volume inbound alone doesn't necessarily paint the whole picture of what "congestion" is happening inside the cordon. Might not work out so well, or maybe it would, it's very hard to tell which is probably why it was not considered due to the complexity. Nothing about measuring congestion in the CBD is simple, trust me on that one, the traffic on the street grid is probably one of the most complex systems in the entire city.