Perhaps someone will be kind enough to correct any errors, but... Here goes.
My understanding is that the fundamental concept of what an IR does is the same as an EQ, i.e. it is increasing or decreasing specific frequencies. However, it is doing so in extremely high detail (I assume dozens of specific frequencies). I also assume it is adding specific reverberations to different frequencies.
The alterations are based on captures of an entire signal chain. Let's hypothesize that a full IR setup modifies 100 different frequencies. You'll have one set of modifications based on:
- the speaker
- the speaker cab
- the specific mic
- the distance of mic from the speaker
- the placement of the mic relative to the center of the speaker
- the reverberation of the room(s) where the IR was generated
- any phase effects of the setup
Even if you had a 100-band EQ, it would be impossible for any human to nail the settings to emulate a specific signal chain.
AFAIK, IRs should always be as close to the end of the signal chain as possible. I happen to put the HPF first, so that low frequencies aren't sent to effects. My guess is you could put the HPF at either the start or the end.
IRs replace cab sims, but are not intended to replace EQs, HPFs, LPF, exciters, and so on. The job of an IR is to replicate a full audio environment when going direct into a flat system (e.g. a PA or DAW). HPF/LPF cleans up your sound so that your amp isn't producing frequencies detrimental to your tone. An EQ is a broad brush that shapes your sound.