Some issues with the new Harvard 1000 Genomes data

After downloading and testing the Reich Lab data labeled 1240K/5081 individuals I found some issues with the included 1000 Genomes data.  It looks like everything else is OK and the issue applies to 1000 Genomes only.   I prepared a replacing data using the original project data and results are fine.   I am very grateful of all new ancient samples, but have to inform about this.  The issue probably doesn't become revealed on PCAs using global data, only in local European analyses.