...

That configuration is optimal for maximizing recovery times and write latency: since each EC set will span 8 hosts of 9, every time a host becames unavailable virtually all PGs will become degraded, and even with all host available any write will involve nearly all hosts.

...

This kind of optimal configuration is popular with those who think that the original design assumption of Ceph of having many, many more hosts and OSDs than the replication factor or stripe width was wasteful and cowardly. :-)

Isn't a bigger k+m a good things for performance if network and hardware not the bottleneck ?

In some sense, but it also matters how many hosts you have, so it's more about "how close to degraded am I when everything is normal?" If you have 6+3 with only 10 hosts or 6+3 with 100 hosts matters a lot if one,two or three hosts accidentally goes down. So it's not only about if "6+3 is better than 4+2", but also how large part of your hosts this covers. -- May the most significant bit of your life be positive.