it seams to be problem with the Statistic tool; it works on some bands but on other the behaviour is not as expected or crashes.
The tool has been used with a ROI and 10m bands:
1.- when selecting for instance B02 the tool works as expected:
Ok, obviously the problem originates from having multiple pixel resolutions in the same data product, as it is the case for S-2 L1C data products. It is a known issue we which is planned to be resolved in the final 2.0 release. Actually this issue is one of the major reasons why we are still releasing beta versions. Sorry for the inconvenience and thanks for reporting and your patience!