of course this is an "idealized" graph, but there are three overlaid curves--with the filters switched off, there's a flat gray line from 90 Hz on down to 20 Hz at -4 dB, which is somewhat hard to see but it's there. Are you saying that you hear a rising response below 90 Hz, or ... ?
And yes, of course, every frequency response graph of a directional microphone is drawn w/r/t some particular measurement distance, contrary to what the IEC standard specifies ("plane wave" conditions, i.e. no proximity effect). Most of the top-rank manufacturers have adjusted their measurements for many decades to describe the behavior at 1 meter as a compromise. For more distant recording, the microphones will have less actual low-frequency output than what the graph indicates. The big exception is DPA, who adjust their measurements for, I believe, either 50 or 30 cm, I forget which at the moment. That throws off any comparison of their curves with anyone else's, making their directional microphones look on paper as if they have fuller, more extended low-frequency response by comparison.
Again, this doesn't refer to their omnidirectional microphones, since proximity effect isn't a factor there; also I'm not going to deny that DPA makes truly excellent microphones; it's only their published response curves that I'm grousing about.
That said, even if you were tipped off that manufacturer X adjusts to 1 meter while manufacturer Y adjusts to 15 or 30 or 16.2 cm, what correction should you apply to put their graphs on an equal footing, or to show the expected performance of their microphones in your actual application (e.g. more distant recording)? There's no formula; it depends on the acoustic design of each microphone (mainly on the transit time for sound waves between the center of the front and the center of the rear of the diaphragm, which is largely but not exclusively a function of the capsule diameter).
In other words there's no substitute for actually measuring under appropriate, highly similar conditions and reporting honestly. But if anything, the industry has been moving for years in the opposite direction from that.
And by the way, the farther you go along the spectrum of directional patterns from omni through cardioid, supercardioid and hypercardioid to figure-8, the more difference the measurement distance makes; proximity effect is considerably greater for a figure-8 than for an otherwise comparable cardioid. The attached graph is from an old brochure for the Neumann U 67 large-diaphragm, three-pattern microphone (released 1960).