A new system just doesn't gain anything in repasting. Apple uses a good paste while some will argue, the amount of improvement putting in the very best paste is 1 or 2 degrees. You'll not get any benefit with so little an improvement.
Think how a race car driver is tuning up his car for the big race. Here he's trying to maximize his car in a car duel with others. Now look at that same person driving to the race course he's not shaving corners or worried about the slow car in front of him. Very different patterns here. This is also true here. This system is not a super computer racing though calculations, its a general purpose laptop designed for portability for general purpose computing.
To your direct question:
There are two issues you are dealing with, the first is making sure you don't create any air bubbles, the second putting on enough so the chip is fully covered and not spilling over.
I personally use the pea method (a dab in the middle so it spreads out evenly in all directions). As you press straight down the paste will spread out pushing the air away.
Spreading a coat of paste only works if you press from the edge tilting it to the other side in a wiping motion, otherwise you'll likely get an air bubble or two (think how you turn a page in a book). Most of the time you don't have room enough to do this and will likely use to much paste.
Air bubbles have two effects they reduce the surface area and prevents the heatsink to be as tight as you can get for better heat to transfer. Remember you are trying to fill the microscopic voids not add extra thickness to the sandwich.
The best way to practice is to get some glass microscope slides and use a few different viscus liquids like honey and gel toothpaste. Here you can see how the way you apply the top glass effects the filling of the honey across. Hint do the end corner L configuration so you can pull the glass apart without breaking. Using different viscosity liquids allows you to see how they react. Some paste is runny others stiffer.