The FT headline "Coronavirus death toll in UK twice as high as official figure" needs some qualification, above all that it is an FT extrapolation of previous stats. Not a report of current stats. This is the basis of the extrapolation:

Using metric "excess deaths from all causes" ("it is widely recognised as best measure of death toll linked to pandemic") and latest ONS data, from mid-March, as Covid death toll mounted, to April 10 (latest ONS stats) "excess deaths" above seasonal average were almost 17,000.

The FT then takes the latest trends in daily hospital deaths (which have now reached over 17,300), assumes the relationship between that metric and "total excess deaths all causes" remains stable and extrapolates what that would mean for total excess deaths now.

That's how it reaches its "conservative estimate" of UK excess deaths of 41,102 by April 21. Over twice the aggregate hospital deaths total by yesterday. We will know if this extrapolation is roughly accurate when we get the ONS figures for w/e April 17 and 24.

The headline involves assuming all excess deaths are Covid-related. FT says it has expert cover for that assumption. British actuaries yesterday estimated excess deaths had reached the region of 28,000 - 34,000; but don't assume all these excess deaths are due to Covid-19.

The FT reports that ONS data show vast majority of all excess deaths were people aged over 75 years old. This age bracket accounted for 70% of the total, the same proportion as those with Covid-19 on their death certificates.
As always, measured comments on this thread welcome