@Midlifephoenix
They are not 'known to be inaccurate'.
On the government website it says they have a 'specifity of 99.9%'. That is pretty darn accurate.
Let's be really clear here:
They are 'specific' - if you get a positive, you are extremely likely to have Covid (though at apopulation level, the false positive rate is an issue, because even at less than 1 in several thousand, with your average secondary doing 3000 tests over 2 weeks, there will be false positives).
They are NOT 'sensitive' - as a rough estimate based on different studies, when done by patients themselves or under the supervision of relatively untrained personnel (as opposed to when carried out by scientists at Porton Down), they detect 50% or so of the infections that would be detected by PCR. The way the government presents this is that the LFT tests detect those with high viral loads, who are most likely to be infectious. However, I don't think that the 'minimum viral load' in person A to be infectious to close contact B has been established, so I understand this is an assertion, not absolutely factual.
What the layman means by 'accurate' is probably the second - does it detect if you have Covid? - or a combination - does it give exactly the right answer, whether positive or negative - every time? The answer to both of those is 'not really'.