I used to leave it out (for more-or-less the reasons you say).  Then last month I decided to put it in in the interests of having all available data to look at, even knowing some of it was less than perfect.  Today, I was debating leaving it in but coloring it differently and underweighting it in the moving average (since there's presumably less information there than in the months where we have a corrected IEA estimate and and EIA estimate too).  However, I decided that if I change the methodology every single time, we'll never be able to compare how things are changing.  So I think I'll leave it this way for a bit.