http://www.dhv.de/web/en/safety/articles-statistics/ltf-a-and-b-class-saftey-check/

The test series conducted here clearly supports the general
suspicions regarding LTF and EN testing: the norms are at best a coarse sieve –
large discrepancies are quickly found, but smaller ones may find their way
through. We haven't found any really dangerous gliders in the A and B classes,
but it is somewhat frightening to see gliders tested to be as safe as possible,
that still require 60m of height to recover from a massive collapse. In doing so
they pitch forward alarmingly, rotate through almost 360° and have sink
velocities of over 20 m/s. This is not the kind of glider that belongs in the
hands of a beginner. Accident investigations clearly show what beginner and
low-airtime pilots tend to do in the event of a collapse – nothing! They are
usually much too frightened and inexperienced to calmly and coordinately react
in the heat of the moment. What they need in such a moment, irrespective of the
piloting errors they may have made beforehand, is a particularly friendly glider
with moderate reactions. This is often promised for the LTF-A class, but these
promises are not always kept.

The performance gliders in the high-end B class are marketed, quite correctly,
as cross-country machines. They belong in the hands of experienced pilots and
are definitely not for “Sunday” flyers. The thin line between “just OK” and
“clearly over-demanding” regarding what pilot skills are required to recover
from instability provide a lot of food-for-thought. A slightly steeper folding
angle on an asymmetric collapse, a bit more of the span on a frontal collapse or
a little more sink in a spiral can change a moderate glider into one that's hard
to recognise again. Between “typical behavior for its class”, and cravats, dives
or stable spirals lies very little margin for error in some cases. Gliders in
this segment are only for pilots capable of active flying, able to recognise the
onset of instability and react immediately to prevent collapsing.

The general impression we were left with is that more intensive testing is
required, and not just two norm flight tests for certification. To provide
realistic judgments on glider characteristics, a test program with several
collapses, stalls and dives is necessary. Only then can we determine the entire
testing results bandwidth and inform pilots appropriately.

We often hear of the evil surprises that some gliders provide in extreme
situations, such as the fatal crash last season of a school pilot after an
asymmetric collapse, cravat and spiral dive into rocks. Or the pilot with 20
years experience who moved up to a high-end B glider and dies after not being
able to exit from a spiral dive.

To simply write off these incidents as “pilot error, bad luck” does not do them
justice. Paragliders are built for pilots, and pilots do make mistakes. In
glider classes for beginners and low-airtime pilots passive safety
characteristics must have utmost priority. And there we still have plenty of
room for improvement.