"Jason Crusan, director of Advanced Exploration Systems at NASA Headquarters, said that the inflation process was stopped because the habitat wasn't expanding as quickly as the computer models predicted."
Good example to illustrate that you need to keep to be aware of the possibility that reality won't comply to your model(s), especially not the untested ones. Which is one of the reasons why stuffTM needs to be tested IRL. Which is a bit complicated if it's stuff-for-spaceTM.
No doubt the team at Bigelow was expecting something to work out differently as planned - now they know what did and will work it into their design and performance models. Other things might work out differently as well, which is why a two year test period makes sense to me.
As I am given to understand that there already is a technology to inflate limp things by exposing them to a localized near-vacuum, that has been around for some time, I was initially puzzled by the failure to expand the module at the first try. But now that I know that it had been in a tight squeeze for 15 months, three times the duration originally planned for, I'm quite sure that this is the root of the problem.