Oh I hate these kind of analyses.
Take the metric used:
The resulting Safety Index relates the accidents to the revenue per passenger kilometre [RPK] performed by the airline over the same time.
Well BA lost a 777. But the accident was a flaw in the 777, so could have happened to anyone, it just happened to be a BA plane. How are BA a less safe airline because of that?
Likewise, the US Hudson incident - how does the fact that a US plane ingested geese make them a less safe airline than an airline which has not ingested geese?
Airplane loss is a crude, meaningless measure. While some airlines (AA, SQ and AF come to mind) have lost planes where crew training or management has been implicated as a cause, which might highlight deficiencies in their safety, other incidents/accidents are the result of a set of coincidences and the airline involved just had the misfortune to be the one involved but did nothing to determine if the accident would have happened.
Indeed, one could argue that the training of the BA and US pilots I mention above stopped fatalities in both those incidents - but that is given no credit and they are dinged as 'unsafe' airlines because they have lost planes.
So really, not worth the paper it's written on. And QF lost a plane in BKK, but didn't they pay an awful lot to get it back into service just so they could continue to state they hadn't lost a plane?