From controlled lightning test I've seen a car body works quite well. It's not a perfect Faraday cage but it works.
Ever watched those guys dress up in chainmail and stand between two mofo Van de Graaff generators? Same principal applies.
It's not really a Faraday cage example. The reason it works is because being surrounded by metal makes this the path of least resistance. Any other path followed inside the car would add to resistance and the lightning would have to then find a path leading back to the frame so it instead avoids this and takes a path through the frame as though it were solid.
Back to the topic: In order for a signal to be blocked there needs to be an obstacle larger than the wavelength with no openings larger than the wavelength. This is not a hard and fast rule though. As said the signal will only be reduced and true blocking only comes in when the obstacle is also thicker and of the right material. In your case though the building probably has a bigger effect on the signal and the window even with bars could help it.