I agree with basically everything you said. However, this would only apply in relatively narrow cases where your workload can be supported by a single RSM. Ther

I agree with basically everything you said. However, this would only apply in relatively narrow cases where your workload can be supported by a single RSM. There is an absolute limit for the sequencing alone from a single leader, even after you apply the tricks mentioned in the paper. Geo optimizations (US users data stored in US, EU users data stored in EU, etc) for latency while maintaining a single database view also become challenging (impossible?) without sharding. Fault tolerance is also slightly improved with many shards because you expose a smaller fraction of your database to transient downtime (re-election, catchup) or slowness when the node containing the leader fails or becomes slow.