wizard71,
That's exactly what I did, trying to eyeball the mic over the spot on the grid as accurately as possible. If there was nothing between me and the mic and the floor, it was pretty easy to do. If there was something, like the chair, in the way, I would look from two directions 90 degrees apart and a few feet away. I've heard of some people using a plumb bob, which could potentially be slightly more accurate. But the chair would get in the way of that for the close in spots too, so I figured I'd just be as consistent and accurate as possible by eye.
 
To make my grid, I cut up shiny return address stickers (everybody must have a few hundred of those lying around that they received unsolicited) into squares and stuck them to the floor according to my measurements. This made it pretty easy to visualize the vertical alignment.
 
It worked for me!

When I setup my ARC 2 system I moved my mixing chair out of the way as that seemed to make sense.  From what I understood of the instructions, you want the mic to be where your head is at least for the first measurement and then you move the mic out to each side equal distances etc. for the other measurements.
 
I made a grid on the floor with a measuring tape and then marked measurement points with a black dot, then I taped a laser pointer to the microphone aiming down.  I used a microphone stand to hold the mic (the manual mentions doing this) and got the stand as out of the way as possible, that is to say the arm of the stand was perpendicular to my speakers (I kept his perpendicular orientation for all the measurements).  I would then move the whole stand etc. and while using the laser I could get the mic into the correct position easily.
 

I did mine the same way except I used a plumb bob as I don't have a laser pointer.

As I recall, ARC lets you run as many set up's as you wish. And you can save them all with unique names.
 
So yeah, do several and do with and without you in the chair. I really don't think the difference will be that much.

I'd remove the chair from the room for ARC, just for convenience while moving the mic stand around. Unless you have one of those giant high-backed overstuffed leather executive office chairs, it won't have a significant effect compared to the room's contribution. (It will, however, skew your perception of some high frequencies, so a plain low-back office chair is preferable at the mix desk.)
 
Your own body is pretty absorbent and will have a greater influence than the chair, so it might seem to make sense to have it be part of the snapshot, but it really won't make that much difference. Your body does absorb sound, but because it's at the same location in the room as your ears it has little effect on how you perceive sound. You're more interested in the effect of objects far enough away to reflect significantly delayed sound waves: the floor, ceiling, walls, and large objects such as couches and bookshelves.

I leave the chair in the room, but move it straight back from the mix position just far enough to be able to move the mic stand around. I then try to get behind the chair during measurements, but that 3 second countdown does not provide enough time to crouch (would really like to know how to increase that). If I try to hurry I sometimes end up bumping the mic stand.
 
To position the mic I use a tape measure to set the distance from the "C" in the Focal logo on the monitors and write the measurements down so I can repeat them with the door opened and closed. Making small adjustments to the mic position this way is tedious, but it works.
 
I also compare the uncorrected ARC 2 measurements to ones taken with Room EQ Wizard to see if they are similar. The size of the grid in ARC 2 exaggerates the peaks and dips somewhat, but is a good indicator of certain problems if they are really far off from the REW results.