Here's a thought experiment. Say we define "sexual assault" as being any case where a person is touched on their skin by another person, without that other person first getting explicit permission from the "victim" person to do that. By that definition, I'd wager that 99%+ of all women alive today have been sexually assaulted -- or if they haven't already, 99% will be.

Now, in this thought experiment, let's start narrowing the requisite conditions. Say, that the area touched has to be near genitals or buttocks or breasts. Bet the percentage drops some, but still high. Let's say we widen it so it can be outside of clothing -- not necessarily on bare skin. Guess what, the percentage goes back up. Now let's specify that the touchee is not sexually attracted to the toucher. Percentage drops. Now we could narrow it further, or widen it, whatever, and we're going to end up with whatever percentage we want -- even one we want to call "high" or "prevalent" or an "epidemic", if we wanted. Let's add in some error margins and uncertainty because (1) people don't always tell the truth, and (2) people don't always remember things perfectly. Now the importance and validity of your percentage becomes, well, let's say arguable and ambiguous, at best.

Have some women been "raped" (in the clearest, most old fashioned sense of the word) -- hell yes. And that's horrible. And should not happen. But to say that a third of all women are "sexually assaulted" in their lifetimes and/or to say that that is (always and necessarily) an evil thing -- that part I'm not so sure of.

Rape? Bad.

Sexual assault? Depends on particulars of the specific case -- not what words/laws say.