YoY change in jobs is still probably not the best way to visualize overall market health. As you say, you also have to take into account the number of people of fill the jobs. To me it seems like the least misleading statistics would be a graph showing unemployment and underemployment % over time. I'd probably also toss in graphs of length of unemployment period as well as various median wage percentiles (quintiles or deciles maybe) over time.

It shows growth or decline but it absolutely does not show what the title implies.