I haven't posted in a while .... Anyway, it is true that the actual number of clocks is pretty small. In fact if you consider the pulse to be an impulse, and send it at time t, then every one will come back at some time t' but you will measure it (in clock ticks). The main problem is that you don't know whether it was close to the beginning of the clock period or the end. There are plenty of tricks to get around this problem such as histogramming, but consider this rather simple approach:
The receiver will assert some signal when the laser pulse gets back. Now even if the receiver only runs with a 100ns period (10MHz). We could get
a very accurate measurement if we knew where within the period the pulse arrived. The idea is that you take the signal and route it to many
flip-flops each one gated by a clock that is slightly delayed with respect to its peer. For instance, Bit 0 is at t0, Bit 1 is at t0 + 5ns etc. You do this up to half the period, and then flip the clock and now you have the full
period covered and now when you latch the bits, some will be 0s and at
some point you will see 1s. Knowing what the delay is, you can then very
accurately determine the pulse measurement.
Another approach is to make the pulse operate at a faster frequency than
the receiver. The pulses are received and accumulated by a histogram.
The mass-center then tells you where the distance was (not exact, but for
speed you only need x'-x) and you do it again and again and the hist. will
"translate" giving you the velocity.