That's the effect of world war II and the collapse of the french government after the nazi invasion. If they excluded years before 1950 it'd look much more normal.
Note that the absolute worst year to FIRE in France was 1943. From 1940-1949, french equities returned -7.6%/year and french bonds returned -21.7%/year (essentially prewar bonds became worthless). That's a bad enough set of returns to make essentially every FIRE date from 1920 to 1945 a failure, and since that is 16 of the 86 total 30 year intervals in the dataset used in the study...
Isn't it also misleading to start the analysis in 1950 in the midst of the Marshall Plan and general post-war recovery in Western Europe? What are the relative performances if we start in 1960 or 1970 instead?
France 1960-2000: 6.4% CAGR stock returns, 4.9% CAGR bond returns
Germany 1960-2000: 5.6% CAGR stock returns, 3.7% CAGR bond returns
USA 1960-2000: 6.8% CAGR stock returns, 2.8% CAGR bond returns
France 1970-2000: 8.3% CAGR stock returns, 6.1% CAGR bond returns
Germany 1970-2000: 6.1% CAGR stock returns, 3.9% CAGR bond returns
USA 1970-2000: 7.2% CAGR stock returns, 4.1% CAGR bond returns
(The 1980s were a very good decade for France, which starts to skew the numbers as we get down to shorter total historical time periods. The 21st century has been very good for Germany, but the book I'm using doesn't have data past 2000.)