yes this is not a linear relation
Here is an example of flow vs pressure curve - for instance the graph for the papst 4212 NGML (38 dBA @12V)

and the 4212 L (28dBA @12V)
And more interesting they have "variofans" so they publish speed vs pressure/flow curves:

this is for a 80x25 variofan, controlled by a thermistor, so it's temperature which varies, but basically the thermistor (with its control circuitry) varies the voltage between 8V and 12V. The only problem is the thermistor response to temperature linear ? Like billbartuska said we are a special population - when engineering products for the industry the provided graphs are enough, you just want to know temperature vs flow, the actual voltage / rpm is not relevant.