This thread reminds me of the dear old bag lady living on the street who was asked why it was that she carried a thermometer in her shopping cart.
"Well," she answered with a snarly smile, "when sleeping on the streets at night, how would I know how cold to get if I did not have a thermometer?"
My collimation tools are all home made. Using a sight-tube, a Cheshire Eyepiece and a Cheshire/sight-tube combo tool, I have purposely put my primary mirror out of collimation to see what effect it would have on the image in the FOV.
I have since determined that, under my skies, with my gear and as viewed through my eyes, there is a remarkably large 'collimation' tolerance zone in which the quality of the image does NOT change at all.
To quantify this. With my Cheshire type tools I'm adjusting the tilt of the primary mirror to move the reflection of the peep-hole into the center of the doughnut. This is more like what a user would see in a well illuminated collimation cap rather than a typical Cheshire/sight-tube with a well illuminated inclined plate.
The doughnut decal is white but appears to be black in the reflections. The peep-hole is blocked by my eye and so appears to be black against the grey colour of the Cheshire's cap. Ergo, a black doughnut and a black peep-hole seen against a well illuminated light grey background.
The peep-hole on all my tools is 1.6mm in diameter. My doughnut has an inner dimension of 3.7mm. I am finding that if the reflection of the peep-hole is moved anywhere inside the inner circle of the doughnut, there is no additional advantage in tweaking the primary.
It matters not if I move it middle-for-diddle. The image does not improve.
If, after setting up and cooling, the reflection of the peep-hole is seen to already be somewhere, anywhere, inside the doughnut, then I do nothing else. If the peep-hole appears to be somewhere behind, or away from the doughnut, I quickly tweak it to somewhere inside and that's enough. Genuine!
Likewise when it came to checking the centring of my doughnut decal on the primary. I figured that if it's centre was within 2.4mm of the actual optical centre of the cheap, mass produced Synta mirror, it was within tolerance and good enough. It was and I left it alone.