Yes but they do tend to net off deaths that would have happened at some stage in the year anyway against actual Covid-19 deaths as measured by actual diagnosis in widespread community testing regimes. Some of the latter  might have died anyway from co-morbidities or age, but it doesn't mean they can be subtracted from Covid-19 stats or that ONS data is directly comparable to actual Covid-19 stats.

Ireland has recently added non-confirmed but probable covid-19 deaths to its Covid stats - in line with international disease diagnostic conventions which has increased its Covid deaths by another 20% on top of its confirmed numbers. With a median age of 83 and with the vast majority having underlying conditions, may of these would not be regarded as "excess morbidity" in ONS type statistical exercises.

In the US deaths are often just recorded as respiratory failure or pneumonia related and not counted in Covid-19 stats unless confirmed by tests, and there are reports of doctors being pressurised not to note Covid on death certs. So there seems to be a concerted effort in US and UK to minimise the pandemic impact probably largely for political and PR reasons.

Some dry statistical report on "excess morbidity" won't have the same impact when released in a years time. Statistical probabilities don't have names and faces and families...

Index of Frank's Diaries