"from a security perspective, I understand the need for better information"
Quality versus quantity. This is about wholesale surveillance -- gathering as much information as possible on as many people as possible. That does not mean the information is better or even useful.
I bet it is very useful. If you have one or two suspicious persons who happen to be a member of a particular lodge, church, mosque, etc. you can probably usefully winnow down the list of associates to further investigate without a large and intrusive effort investigating the entire group.
People easily imagine the worst case scenarios, but the okay case scenario is surely more typical, e.g. when a man travels to eastern Pakistan to allegedly visit relatives for several months, the FBI can look at the 9 people he spends time with rather than creating dossiers on the 112 people who so happen to attend the same mosque.
(Whether and how the Constitution cares about the "good case", "okay case", "worst case" is a different question.)
The problem is that the scale messages makes even seemingly small false positive rates a big problem. Suppose you had a false positives only 0.01% of the time -- a ludicrously generous estimate. When you are looking at hundreds of millions or even billions of messages per day, that is far too much noise for the system to be useful.
Maybe the FBI/NSA is so good that they never see false positives. I doubt it, though, given how many non-terrorists have been harassed at airports, had their peaceful protest groups infiltrated by undercover agents, etc.
Quality versus quantity. This is about wholesale surveillance -- gathering as much information as possible on as many people as possible. That does not mean the information is better or even useful.