Cotter DJ, et al. (2004) Hematocrit was not validated as a surrogate endpoint for survival amoung epoetin-treated hemodialysis patients. Journal of Clinical Epidemiology 57:1086-1095, Figure 2

What's wrong with this one?

The perspective makes it difficult to compare the heights of the cylinders, as the vertical scale changes from front to back. Also, this is a lot of space (and color ink) to convey very little information.

What should have been done?

It must be admitted that it's tricky to fix this figure; one might try four superposed lines.