It's bad reporting, but the probably don't waant to use time epxlaining why, and likely not many would understand it anyway, or even care.
From the modelling, and S-Gene surveillance, Omicron is 90% of London, and over 50% for the UK now, these are the numbers to follow, and estimate from there. Also factor in for the 3-4x of infections which are missed.
Take the "Omicron Cases number" and throw it in the bin, they shouldn't even reporti it, as it's massively misleading, as that only represents the Covid infections which are then tested (around 1/4th to 1/5th), and an infection only becomes a case after a test of course. Then a small number of these tests are then sequenced, and confirmed Omicron. But, as we only sequence a small number of PCR's, and zero LFT's, that Omicron figure will always be way lower.
If we sequenced every PCR and it was found as Omicron, but we had twice as many user reported LFT's, then the max Omicron cases confirmed would be 33%.