Lamarr worked within the constraints of the technology of her time. The idea was to use the code to rapidly change the transmitter's channels in lockstep with a receiver. A simple way to comprehend this idea is to imagine a simplified 2-bit piano roll representing channel 1 and channel 2.

Table 1. Imagine the data stream being alternately switched between channel 1 and channel 2 over time.

An eavesdropper scanning the radio spectrum would detect such a transmission as looking like this, as the transmitter alternates between the two channels.

As you can see, 2 channel spread spectrum is not so good, but bear with me through the example.

More Channels = Harder To Detect

And now imagine, if the C/A code sequenced through four channels, it would look something like this:

Given the same amount of energy, the signal spreads out more and has a lower profile.

If you take this to the logical extreme, the bars will become so low that they touch the noise floor and seemingly vanish.