Not only that. Access time only determines the latency of one single I/O request; and HDDs can only process one at a time. But SSDs can process multiple I/O at a time.
So with 1000 files and 0.01ms and 64 queued I/O's that would mean that in 0.01ms you would be able to process 64 I/O's instead of 1. That would lead to:
(1000/64) * 0.01ms = 0.15ms.
Of course, this would be the most extreme scenario where the SSD is superior to HDD. We talked about I/O's with 0.01ms access time, but that does not include the transfer time. Also, current SATA 3Gbps and Windows OS is not able to saturate the SSDs in terms of enough I/O commands queued. So lots of reasons this scenario doesn't come true in regular scenario's.
Still, it shows a fundamental change: SSDs are parallel I/O devices, while HDDs are serial I/O devices - able to only do one thing at a time, at least for the mechanical part.