It's more along the lines of using that metadata to know what time you visited a website so they can match activity on the website to your timestamp. This is also how they have in the past found you're responsible for something without having to get logs from your VPN. Another usage where this will be useful will be for peer to peer networks, now police can easily find out exactly what you have downloaded via bittorrent. And if tox.im type of 'p2p chat networks' become used, that would help them more than it would if you used a centralized chat service as it's previously been stated that the metadata (who communicates with who) is more useful than the contents.
On another note, do we know yet whether the police could bulk ask for everyone's connection history or do they need 'reasonable suspicion of a crime'?
As far as I can tell, bulk collection does require a warrant (as per the bulk powers section) and is limited to security / intelligence agencies (does that include police?).
Police have to request the ICRs for an individual on a case-by-case basis, going through a 'senior officer' who takes advice from a single point of contact (SPoC), although I can't find any criteria for SPoC selection.
In any case, there's no judicial oversight on ICR requests as far as I can see.
On another note, do we know yet whether the police could bulk ask for everyone's connection history or do they need 'reasonable suspicion of a crime'?