Well, that would be the exact question that the sabermetric evidence examines.
I guess the high K%, high HR% players are just more often neutralized by higher level pitchers. Something like that.
Like, these elite SP are giving up like 0.5 homers per nine innings. Elite relievers, even less (some of them go all season with giving up one or no homers).
Over a big enough sample, yes the homers will happen and they have big impact but in a short series the ability to put the ball in play and get lucky can help tip the odds in your favour. A little bit. Just a little bit, probably.