@Aristotle's Horse, you're absolutely right.
in principle, there is no problem in the police and security services having the ability to monitor those who are suspected of a crime, subject the approval and oversight of the judiciary. This oversight is a vital safeguard in a true democracy.
The involvement of the judicial branch is a measure to prevent abuse of those powers by 'bad actors', within the police or government. it is not reasonable to presume that without such safeguards, these powers will always be used correctly, and history would indicate that they will be abused.
Even though the involvement of the courts would most likely be a 'rubber-stamp', it would mean a record of the request is kept (although this may remain secret for a number of years as national security requires). Furthermore, such 'rubber-stamping' is absolutely not the job of a politician (i.e. the Home Secretary), as there is a clear conflict of interest in cases where investigations may have a political element.
Note that the existence of such powers is in itself not a problem, if they are used correctly. On the one hand, we have the fundamental human right to privacy, but on the other, we have the rule of law, and the need to prevent criminality. This leads to the sensible conclusion that intrusive surveillance should be used only to combat crime (which includes terrorism), and not to watch the populus as a whole. This should require reasonable suspicion, so that it is not applied inappropriately, limitations so that it is not over-reaching, and oversight, so that the results can be evaluated independently to establish whether the surveillance was justified in the first place, to check any possible pattern of abuse. None of these prerequisites needs to be onerous - as I said, it is a simple matter of rubber-stamping and recording.