Using the French national statistical service's tally of deaths,, comparing to the announced number of hospital deaths of people who tested positive for COVID, it looks like that number represents about half the death rate, at 10th April. The official number which includes rest home deaths looks like it's about 25% short of the real number which we will know eventually.

The first graph gets the peak in the daily toll in roughly the right place compared to government figures.

Bearing in mind that the difference in daily mortality is mostly about seasonal influenza at this time of the year, and that this year's flu is fairly mild, we can compare to 2019 and be within a few hundred when estimating the excess deaths from COVID. The most recent data cover about 90% of the population (the communes that file death reports electronically, rather than through the post).
Annual deaths so far by year :

10 April 2020 : 79665
10 April 2019 : 64210
10 April 2018 : 70997

So, inflating by 10% (though that may be excessive, since the communes mailing it in will be smaller rural ones, with low death rates), 16 or 17 thousand extra deaths at that date,  compared to 8598 hospital COVID deaths, or a government number of 13197 total COVID deaths to 10th April.

So the official French numbers look like less of an undercount than I expected.


_{It is rightly acknowledged that people of faith have no monopoly of virtue - Queen Elizabeth II}