There are some nasty potential error sources when comparing productivity per country over time because they use different currencies with their own value changes over time. (Probably related to productivity, but lots of other factors)
And my biggest issue was the Wells Fargo twitter chart. Average growth per year is an ok tool, but it's simplistic. It gets really messed up when you consider a 5 year time frame 98-03 and a 10 year 04-14. With the straight averaging method compounded growth/loss gets messed up.