I interviewed John Moller, the man responsible for the Unpubs, while I was there on Saturday. The video is below. Also, here are some photos of the Unpub.
Update: Here is another summary of the event.
At the event, feedback forms were distributed to all playtesters after each game. If you are curious, you can check out the feedback received for Municipality. I was quite pleased with the results. Most of the players had positive feedback.
In addition to the forms, I made sure (as I always do) to verbally ask the players about the game. This managed to get more negative feedback out of them. By "negative", I mean that the players identified weak points in the game. In particular, one of the three tests of the day managed to finally convince me to change something I had foolishly thought was core to Municipality. So, yes, I am eliminating the split between adjacency and connectedness. From now on, the only thing that will matter to a building is what buildings are adjacent to it. I will adjust the effect numbers so that the overall levels seen will remain about the same, but players will no longer have to trace routes on roads across tiles to determine the value of buildings. Consequently, they also won't have to worry about orientation of the tile when placing the building. Dan Cassar was right: the influence-trading part of the game is strong enough that it doesn't need a highly complex city-building game on the front end. This change will require more testing, but I am not short on time. I have plenty of other things to take care of in parallel, like hiring an artist.