I dont think its clear why would more plates be slower. More plates means more heads, with possibility for concurrency - that should increase sequential transfer speed.
If data is written cylinder-wise, the latency should be similar to one-plate disk.
If it is written plate-wise, the latency should vary up and down in relation to block numbers. Its possible the average latency would still be comparable.
The only clear negative about multi-platter systems is the increased inertia of the head assembly. It's not so clear whether it has a practical implication.
Apart from this unclear performance implication, there is of course the decreased reliability and increased cost of multi-platter solutions. That is the main reason we don't see that many of them.