[1] The theory of human evolution we were taught (who knows, maybe it's still being taught that way...) dates from when the "bushiness" of the family tree was unknown; also, before anyone gave credit to catastrophic causes of extinction.

[2] No 'gradualist' evolutionary model works in the real world - [a] the ultimate source of evolution is random point mutations on individual genes, filtered through natural selection; [b] extinction can occur without reference to the fitness of the organism (as a longstanding fan of S.J. Gould, I'm in complete agreement). That goes for entire phyla, not just a single genus like Homo or an order like Hominidae. See Gould's Wonderful Life: you cannot expect to "rewind the tape" (his phrase) and get the same playback.

[3] Regarding the Fermi Paradox (and its cousin, the Drake Equation): Odds against two or more sentient species evolving within eventual "reach" of each other are a lot worse than once thought by some. There are no guarantees - sometimes bad shit happens, sometimes good shit doesn't.