I wonder if there's not an alternative hypothesis for the data from which these conclusions are drawn.
It might not be that "most things don't matter", but rather that everything matters, to the extent that any attempt to measure the effect of any one event on happiness will be defeated. Any event you wish to measure is just one small "nudge" in an endless ocean of nudges that effect you day after day after day. But they all matter, and add up to a kind of random walk of happiness.
How could you distinguish between these hypotheses?