The true magnitude isn't the point, the relative change is, and limiting the graph to a specific sub-range makes it easier to read because it is is more straight forward to following the tick lines to the actual label numbers on the Y-axis. There are a lot of ways to misrepresent data, and data on graphs, but this isn't one of them. These graphs are especially nice because the Y-axis labels are in a largish font, so it's more difficult to misinterpret the scale.

I do have a minor problem with the last graph though, "Monthly supply of homes for sale". Since this is meant to compare existing homes to new homes, the line graph and the thickness of the line doesn't convey enough of the target information; this would be better done as a bar graph, with two bars, one for existing and one for new, for each year.

I had a boss once who made graphs with three Y-axes on them (two on the left and one on the right) and multiple X-axes, all in different scales, mixed linear and exponential. He tried to use different colors, but would then print it out in greyscale. They were extremely hard to read and get anything useful out of, which I always thought was part of his goal.