Roy Spencer has taken Craig Loehle’s 2000 year temperature reconstruction to 1995 and added the HadCRUT3 to 2007. Obviously, the proxy temperature reconstructions during the Medieval Warm Period would have larger error bars than the current (instrumental) temperatures, so one shouldn’t put too much emphasis on small differences between the current peak and the MWP peaks.
Roy Spencer also looked at what the HadCRUT3 trace would look like if temperatures now remained constant for another 15 years (until 2022)…in that case, the temperature trace almost reaches the +0.6 deg C line (just barely
exceeding the warmest MWP peak).
We can clearly see that the coldest part of the Little Ice Age was unprecedented in the past 2000 years, and the subsequent recovery to the Modern Warm Period.