The PLL would fail to lock at random frequencies and temperatures. You'd try to scroll through e.g. the 2.4 GHz ISM band and would randomly get downconverted signal or baseband noise on every other click. Two minutes later, the set of bands you could capture would be entirely different. There was no consistency whatsoever.
When I went to their forums to see if other people had encountered the same issue, it turned out that basically none of the LimeSDRs even pass their own self tests [1]. I don't know how they're still selling these devices... they don't work.
I returned it, paid ~4x more for a USRP, and could not be happier. The hobby is playing with radio modulation schemes, not debugging a defective piece of silicon.
[1] https://discourse.myriadrf.org/t/new-limesdr-mini-loopback-t...