AFAIK, most of those method generate more or less the same result, the temperature difference usually just 1-2 degree C, well within normal error. So, I will say it doesn't really matter. In fact, if you use AS5, their website has recommendation to use which method for a particular CPU model. Or if you go for the Mac Pro's technician manual, there is also a recommended method. Interestingly, none of your 6 pictures including those recommendations from AS5 or Apple.
I personally go for the Apple method (like a stop sign, but turn 90 degree), and put the heatsink on. After that I will remove the heatsink straight away to check if the paste is enough to cover the whole surface. If not, apply a little more on the dry area, then install the heatsink (if obviously too much, remove some paste). I know some people may say that's the worst I can do, because I make lots of bubbles inside the paste etc. But from my real world experience, it works very well. What I found is that the paste amount is the most important factor, but not the application method. May be AS5 is not that sensitive to application method. As long as I use this method to ensure the thermal paste just able to cover the entire surface, I always have good result (that's including my cMP, graphic card, PS4...)