Metadata IS data
The only way to frame this conversation is to list EXACTLY what is being collected, how long it is being kept, who has access and under what scrutiny.
The whole metadata/data argument is a semantic diversion. The reason it is becoming such a farce is that the government has pitched this new grab as not being at all intrusive because it is only metadata. Thus, they must maintain that metadata (as opposed to 'real' data) is a benign and routine thing to keep and certainly nothing for anyone to get worked-up about.
To the semantics, however, metadata is 'data about data' or, data that describes other data. From this it is clear that it is data itself, just a subset, bounded by only definition.
For me, the delineation is primarily dependent on the use/function of the 'actual' data - i.e. what we are interested in. Metadata, by this definition, is generated/recorded for the purpose of identifying the data you are interested in, usually for search purposes
In the case of phone calls, the interesting part, in general, is what is said over the phone. Thus, who you said it to, when, and for how long, can be considered to be data that describes the interesting data. If you were looking at your phone bill, you would use that data to identify the calls you made. Likewise with e-mails - you would use the sender/recipient and sent/received date and time to locate and e-mail in your client.
The important part is that what you are interested in is the 'data' and the details you use to find that data is the metadata. Thus, what is considered metadata is entirely dependent on the use you are putting it to.
This matters greatly when you talk about Internet browsing histories because the URL can essentially be used to reconstruct what you were viewing. If you know the URL, you can, in general, use that to go to the same page and see exactly what the other person saw. You know what videos they watched on youtube and what images they opened on <insert embarrasing pornography site here>. In other words, you have access to all the 'data' about their internet browsing.
Therefore a list of URLs accessed most definitely counts as 'data'. It doesn't matter that if the HTTP requests and responses are stored or not - the URL contains 99% of what you need to know for 99% of web browsing.
More generally, however, as metadata is data about data and can therefore only be defined in reference to the main data it is describing, if you are only storing 'metadata', then at that point, that becomes the 'data'.
But, again, it's all semantic and a distraction. We need to know exactly what, exactly how, and exactly who. That they have so much trouble explaining this clearly to the public is indicative of:
- Blind acceptance of what they're told by the police/ASIO/the US
- Unwillingness to engage in a real conversation with the public
Or, given the number of people who have failed to explain this clearly, all of the above. Whichever it is, it's not good.