This is very interesting! I'd argue that PPP is a very poor way of measuring things like total GDP, but when you are looking at a metric like median household income then it becomes very relevant.
One other factor that occured to me that surely impacts this: this is using median HOUSEHOLD income, not just median income, and different states have different average size households. For example, Maine, which shows up here, has the 4rd-smallest-average-household size in the USA: 2.4 people. Utah has the largest average household size at 3.13. Also, surely, in some states it's more likely that in households with a married couple both are working, and in other states it would be more likely that one was staying at home.