Yeah I know what you mean, according to autodesk's description of stickiness, it is an adhesion force in the normals direction. If that were the case however, 1 should be enough to adhere to anything... it says to use 1 friction (for tangential forces) and 1 stickiness for normal forces to get the best stick... although after playing around with it, it only seems to apply that logic more and more as collide strength lowers....
In fact, the best result I've obtained so far is to actually put an expression into collide strength making it impossibly low, and then increasing iterations. I'd also recommend turning off self collisions and finding another method to solve for that if it's neccesary, depending on your application (like self crossover push or pressure). Another thing that helps is the collide last threshold... I'm not sure why this should ever be 'off'. The only other thing I can suggest is that, if you have a higher poly count try to stay away from Push Out... like if you're messing with the values and never getting the result you want, just set them to 0 and move on. Crossover push and Self crossover push can be useful, but again if you're going for a perfect solve these should probably be 0.
Sorry I can't help further if you're still stuck. Good luck, maybe someone else can provide more info.
The major things that got the best effect in a small test I did:
Impossibly low collision strength (even as low as 0.00001 with an expression)
Quality settings -> Collide Last Threshold to 1.000
Trapped check, Self trapped check
0 push out 0 pushout rad 0 crossover 0 self crossover
1 friction 1 stickiness
Optional, but gives more iterations to the collision and adherement - Turn off self collisions (again, depends on yur application) and decide if you want/need pressure and/or a small amount of self crossover push.
Honestly I have no idea why collide strength would effect adherement at all, but it does, dramatically.