I too have had very good results with the Libre. I compared the results against my Accu-Chek Combo meter and they were consistently within the error tolerances of both machines.
I guess it is possible that some of the posters on the other forum are expecting too much. The permitted error of patient-use BG meters is plus or minus 20% of the 'true' value. So if, at any particular moment, your true BG amount was 10, a meter reading anywhere in the range of 8-12 would be acceptable (at least to the manufacturer). So you could have the situation that the Libre was reading 8 and the conventional finger-stick meter was reading 12 and that would be OK.
In practice, I found that the two machines were much closer than that - which is consistent with the current expectation that BG machines generally work with an error range of 10-15%. I haven't done a statistical analysis of the results I recorded but a rough review suggested that they were generally within less than 10% of each other. Interestingly, I found that the Libre results tended to be consistently slightly lower than those from the Combo meter, but I think the Combo has a bit of a reputation for reading a bit on the high side.
However, I will still use the finger-stick meter to calculate meal and correction boluses. (Not least because the Combo meter does it for me.) What the Libre is very useful for is for checking basals and rapid changes. And, of course you can check your BG through your clothes in an instant.