I do agree with your overall method here but i see one problem. the problem is that every time there is an update from the timeline you are getting the actor location this could lead to problems toward the end of the timeline, essentially if im not mistaken it could lead to an exponential increase of the movement speed of the object being moved. in general it will move much faster than it should. to combat this problem set two stationary points for the lerp. so when the input is pressed set both the start and endpoint as variables.
Edit: Sardorian updated his picture so my comment can now be ignored.