I agree -- anyone who thinks that software projects can be managed only on "numeric factors" without understanding how software is built will probably not do a particularly good job managing software projects. And I also agree that there are a lot of unexpected situations that pop up partway through the project and require that you change the plan.
The funny thing is that the PMBOK(r) Guide -- the framework which the PMP exam material is based on -- says exactly that.
The PMBOK(r) Guide has an enormous amount of information on managing project changes and unexpected situations: what changes need to happen, how to evaluate each change's impact against its benefits, how to make sure that everyone on the project who needs to be involved is given a chance to give input, and most importantly, how to anticipate, influence and prevent unnecessary changes.
It also explicitly says that there is no ideal way to run a project. It says there's more than one way to run a project, and that the only way to figure out how to manage your particular project is to understand the industry and the specific areas of expertise, as well as the project's complexity, risk, size, deadline, etc. This is stated very clearly, and you need to understand it in order to study for the PMP exam.
(Of course, the particular PMP certified people you mentioned may not have gotten those particular questions right!
)
Andrew Stellman
Author of Head First Agile, Learning Agile, Beautiful Teams, Head First C#, Head First PMP, and Applied Software Project Management (O'Reilly)