With a swept wing, you have the same free-stream velocity encountering the wing at an angle. That means only one component of the velocity (the chordwise component) is accelerated, and that component is initially slower than on a corresponding straight wing due to the angle. That would mean that the wing can encounter a higher free stream velocity (and therefore Mach number) before it is accelerated to sonic speed at some point on the wing.
Spanwise flow is both good and bad. As previously stated, it can help raise the critical Mach number, but it can also lead to boundary-layer breakdown and transition to turbulence. On a swept wing (or any body with a pressure gradient that isn't aligned with the free stream) you end up with a the spanwise acceleration of the boundary layer. The resulting 3-D boundary layer develops inflection points which lead to very strong instabilities and becomes the dominant mode of laminar-turbulent transition on most swept-wing aircraft. Right now, this effect can't be mitigated, and aircraft designers just have to live with it and the associated drag penalties in order to reap the benefits of the swept wing. Controlling transition on any wing is one of the holy grails of aerodynamics.