Sounds good to me.
A deployed antimissile weapons system with one chance in five of stopping a missile is infinitely better than one sitting on a test stand in Nevada or on the drawing board.
I see no problem at all with refining and improving it while it's on the job.
