I'm not talking about Cassandra being more generally scalable than other systems, only that no system as scalable as Cassandra offers strict serializable transactions besides Spanner (including Cassandra, today)
The cluster size limitations are also greatly overstated and mostly a hold over from stale documentation from 2013. You can (and companies do) run FDB clusters with 100s or 1000s of nodes without issue.
Cassandra supports this too but in my experience it gets quite difficult to operate at that point, although I’m sure recent versions have gotten much better about this.
If you only care about raw write throughout I think your statement is mostly accurate, but for general purpose workloads I would argue that FDB is a system “as scalable as Cassandra” that also offers strictly serializable (and interactive!) transactions which makes it much more broadly applicable to a wide variety of use cases.