I have been analyzing the data of the three points: 2004, 2009, 2010 that were measured on June 14 and 15. Point 2010 locates at open sky, see please attached Google Map picture, and has 100% RTK fix on both the days. Point 2004 locates at a canopy and has poor fix on both the days, most probably because of the canopy. The most interesting is point 2009 because of big difference in performance of RTK: 15-th is much better, see attached. I am still in process of investigation and going to provide a report when be ready.
As for affect of Galileo, it does not seem to be probable, at least we did not notice such an effect, but maybe we need to run real time processing with reset of RTK permanently and to log 24-hour data permanently to post-process. Periodically, quite rarely, AltBoc signal might be wrongly locked by one or two satellites, and when SNR is low, but this can be treated by reset of tracking or switching AltBoc off.
If total number of signals in an engine does not reach the limit, about 35, it may happen if RTK itself throws out the signals from processing, but this is seen on the screen, or because of insufficient number of tracked satellites on rover or coming base's corrections, but all these are also seen on the screens.