There’s a big degree of “it depends” in this question. That is, it depends on what you’re trying to accomplish on any given track, in terms of how you would get the sound on that track – assuming here you are trying to emulate an analog signal flow as if you were using the actual hardware being emulated by the channel strip plugin.
For example, let’s say you’re tracking a virtual electric guitar (or a real electric guitar that is coming into your audio interface via a DI), where you essentially just have the signal that comes out of the electric guitar, but what you really want is to be tracking an electric guitar that was played through a pedal board, amp, and speakers. In such a case, you’d generally put any guitar pedal board and amp simulation plugins ahead of the channel strip plugin.
Another case might be if you’re wanting to simulate tracking to tape. Would you be going through a console on the way to tape? If so, you might have the channel strip before the tape emulation plugin, but you might also want a channel strip after it (with different settings, the first instance being only what you want to go to tape, then the second being for post-tape, mix settings). If not, then the channel strip would be after the tape emulation plugin. (Another case might be if you’re trying to track through an outboard preamp emulator directly to tape, maybe tracking through that “to tape”, where you might then have the preamp plugin first, then the tape emulation plugin, then the channel strip.)
Of course, there may also be plugins you’d want after channel strip processing in any scenario. I tend to think of channel strip plugins as getting some basics down for the mix processing of a track, but then doing the more creating effects from there, be it still in inserts or via sends to an effects bus.
In any of these cases, though, at least for my needs, the channel strip would be pre-fader. (Not sure if you were asking about that consideration in your “pre or post position” note.)
As for the Cubase channel strip, I don’t tend to use it much, but I think its position in the signal flow may be able to be altered. Where I’ve have used it, though, is at the end, specifically because it doesn’t get frozen when freezing inserts on a track (and I frequently need to freeze tracks with inserts while mixing due to CPU considerations). It’s nice to be able to make minor EQ adjustments at the mix level (still pre-fader) without having to go back and unfreeze tracks and their inserts to make those sorts of tweaks at the plugin level. I mostly end up using that to deal with issues I’ve found after car listening tests or other early “final mix candidate” listening tests that turn up some track (or submix bus) EQ considerations where tracks may be stepping on each other a bit.