While in the case of 1-1+1... the series does not converge to 1/2, or anything for that matter, the value of 1/2 "makes sense" in that it's the expected value of the sum, if one were to stop summing at an arbitrary, random place. This is true for any sum that alternates between increasing and decreasing, but where the "midpoint" between the minima and maxima doesn't "drift" (like it does for, as an example, 1-2+3-4).
However, it makes no sense at all to me to assign a sum to a series that uniformly increases, and by an increasing amount with each term. Even if you can fit a continuous function to the discrete "jumps", that function is certain to approach infinity, so there's no help there.
As someone said, if there is a series or formula in physics that diverges, it means the physics is incomplete. For instance, the existence of quantum mechanics was inferred by Planck on the basis that a collection of oscillating charged particles in thermal equilibrium would, by classical mechanics, emit an infinite amount of electromagnetic energy.