Any microphone can only pick up sound from the sound field where it is. That's all it can do; it can't respond to whatever may be going on elsewhere. There's this mistaken idea that you "aim" a microphone at some sound source/sources a distance away, and it will respond to whatever's there; that's fine for cameras and telescopes since light travels in straight lines. But it's not a good way to think about sound and microphones in most practical situations, since sound spreads out in all directions from its source, and bounces around way more than you might think.
If you record in an anechoic chamber, in free air space far away from any sound-reflecting objects, or in the immediate area of a sound source, then that's a "free" sound field. In those rare cases the straight line/optical metaphor more or less applies, and the MK 2 (which has flat, i.e. accurate, response on axis) is the capsule for you.
Free sound fields don't occur in audience locations in concert halls or other public performance venues, however. At points that are a surprisingly short distance from the sound sources in three-dimensional space, the reflected sound energy of a space adds up to roughly the same net amount as the direct sound. (That distance is called the "reverberation radius".) Schoeps' MK 2 H and MK 2 S capsules are designed for placement around there, since their response averages out to be flat when the prevailing sound field is a rough balance like that.
Those two capsules outsell the MK 2 by a considerable margin; this "semi-distant" type of recording is a common application for omnidirectional microphones--plus if you use either capsule type for relatively close-up recording (as in a studio--German studios use omnidirectional microphones more than American studios do, but they don't always mike as closely), the few dB of gentle high-frequency boost is rather pleasant sounding.
When your microphones are out in the audience area, the acoustical energy at their position is composed mainly of sound that has already undergone reflection off of various room surfaces and people and objects in the room. Thus your microphones are in a mainly "diffuse" sound field, in which the directions of arrival of various sound components will tend toward being random. And each original sound component will arrive along a complex set of paths of different lengths, so the relative times of arrival will be spread over some interval, rather than being "one-time" and sharply defined. In addition, all that bouncing around tends to absorb some of the high frequency content, though the amount and kind of this absorption depends greatly on the materials and people involved.
When an omnidirectional microphone with a "diffuse field" capsule such as the Schoeps MK 2 XS (known as the "MK 3" until recently) is placed in a diffuse sound field, the ordinary kind of published frequency response curve becomes largely irrelevant, since by industry-wide agreement, such graphs show the response that would occur in a free sound field. A free-field graph for a diffuse-field omni will show a high-frequency, on-axis rise--but you won't hear that rise unless you misuse the microphone by placing it in a free sound field or something like it. In a mainly diffuse sound field, only a small fraction of sound energy will arrive within the narrow range of angles where that high-frequency rise is shown on the graph--and among that small fraction, the earliest-arriving sound components will be the direct and "most nearly direct" sound (assuming that the engineer is awake and has oriented the microphones generally toward the sound sources). So that characteristic helps give a clearer impression of the direct sound, such as it is at that distance, while the overall response of the capsule or microphone is essentially flat (i.e. accurate) given the type of sound field that it is placed in.
None of this discussion makes much sense if you use a camera or telescope (optical/visual) metaphor for the way you think about microphones. Instead, consider where the microphone is, and realize that at best, it can only record what's going on where it is.
--best regards