@TheSunIsStillShining What do you mean by reliability, the tests are as reliable as they are (detect maybe 50% asymptotically as they're designed to be used, but significant error bars around) so they're pretty unreliable. But that's different to if they are useful for reducing the spread of cases.
For example if they catch 100% of super spreaders (not unreasonable hypothesis as they appear to be more reliable when there's lots of virus which I think is a reasonable assumption with super spreaders) then even being only 5 or 10% catching any case would likely be very useful, let alone the probably 50%.
Should you rely on the result of an LFD after significant contact with known cases and visiting unvaccinated high risk individuals - absolutely not - but that's not what they're used for.
When cases are low, false positives are high (at least a 1/3rd of positives during march/april were false) too high for me to think 100% isolation on a positive (better to allow solo/family outdoor exercise for example as that has significant health benefits that mitigate the damage by isolation)
I also think they're expensive for the return on health you get for them - money could be better spent on other health interventions, but that's not a COVID specific reliability question.