The AO-7 may correct for mount tracking errors and some seeing induced errors.
periodic error and sudden jerky movements in tracking, are better coped with by the AO-7 than the mount guiding as it is a high frequency corrector:
The brighter your guide star, and therefore the faster your corrections, the more effective
it is.
The AO-7 is truly helpful when it can guide at frequencies higher than 1Hz, and at 2-3Hz, can remove effectively mount problems such as bad mount tracking, PE, or random errors
Past a certain point, around 10Hz, the AO-7 starts making effective atmospheric seeing corrections as well (but see below for qualifications to this).
Even the best mounts can't correct at 20Hz or 30Hz like the AO-7 can. It will almost always be an improvement.
If the AO-7 cannot guide at these frequencies > 1Hz, then it is usually better to turn it off and leave the guiding to the mount.
many find the AO-7 is indispensable for deep sky imaging, especially with mounts such as the LX200 or for focal lengths > 2000mm on a Losmandy G-11. The benefit may not be so noticeable when using the better Paramount ME mount with periodic error correction & software enhanced tracking with excellent seeing conditions, although the actual percentage improvement in use of the AO-7 actually improves with better seeing (you may get 10% improvement at 3 arc second seeing and 25% improvement at 2 arc second seeing).
it may help compensate for atmospheric seeing if the guide star is bright enough to allow at least 10Hz corrections AND it is very close to the imaging field, ie. within the effective isokinetic patch (EIP):
the EIP is the angular size of a patch of sky that appears to move together as one unit due to the effects of seeing & is determined by the system's ability to resolve differential motion due to seeing (ie. the magnitude of the seeing, the size of the nearest convection cells relative to the aperture of the instrument,hysteresis, phase lag, and other factors) and may be many arcminutes in amateur equipment with apertures < 24“ diameter, much larger than the true isokinetic patch which is often ~0.5 arc minute.
As your equipment and conditions improve, the EIP gets smaller relative to the true isokinetic patch, and thus the correlation between what the guide chip sees and the imaging chip sees gets smaller. But it never goes to zero for any instruments that we are dealing with in this discussion. At worst, you are integrating the seeing effects across the aperture (roughly speaking, you are averaging the seeing vectors into a single vector for which you correct) so that there is typically some seeing benefit even for off-axis guiding.
for many amateur instruments in the 1500-3000mm focal length, the AO-7 provides some level of useful correction for seeing _if the correction rate is fast enough_. Again, based on empirical tests, the rate of correction needs to be at least 10Hz, and improvement occurs beyond this minimum. Getting 20Hz corrections will correct more for seeing effects than 10Hz corrections will.
images taken with an AO7 normally show improvement, especially when the seeing is good. The answer is in the nature of guiding corrections. Most scopes, even with quite good mounts, have a 'minimum' step size that the image can be moved by a correction. In the case of a G11 for example, it is around 0.5 arc seconds. The AO7 conversely, can move the image typically by amounts below 0.01 arc seconds. So the improvements seen using the AO7, are not a result of it actually compensating for atmospheric movements, but the result of it raising the correction accuracy by perhaps as much as a factor of 50* (though this is limited by the accuracy of the centroid calculations, and perhaps 10* is more 'honest'). The correction is also fast, and this sort of speed cannot be matched by the higher masses involved in the mount. If you look at more upmarket mounts, the movement steps on the mount are significantly smaller (for example, the AP900/1200, manage a step size of 0.05 arc seconds), and the guiding gain is only from the speed advantage, rather than the improved movement accuracy, and the result is a much smaller improvement. The AO7, is a great unit, for trying to image at longer focal lengths on most mounts, provided there is a reasonably bright guide star available, but on some higher quality mounts, may not create the same improvement.
you will get better guide performance with the AO-7 than with conventional guiding - most programs have optimizations for getting more out of a dim guide star. When your seeing is good, you can use a very small guide box and that helps cut down on the image-to-image timing. But in the end your guide speed is only as good as the available guide stars. The larger your chip, the more you can move the scope around to find that elusive bright guide star. Galaxy season is the worst; there are many fewer stars outside the galactic plane. Some objects just aren't going to work; you have to let those go and image the ones that will work.