An experienced programmer might look at the problem of whether or not the company should replace php 5.4 and determine whether or not it's a completely or partially a good idea based on the factors they have control over and have budget to solve. An in-experienced programmer would probably not have learned PHP 5.4, and so would need to replace it in order to be productive, regardless of accounting for how much value they actually add by doing so.

The end result would probably be well argued strategy of incrementally adopting new ideas, testing them, and re-writing when time permits.