I know this is a very old thread, but I just wanted to correct some issues here as aluminum has become much more apparent to people as they get the data from Triton tests and many people now see it, when I was the only person to spend the time and effort to detect it years ago.
Yes, the first test used a substantial amount of Phosguard. At Seachem's request at the time (many years ago now) I repeated the test at exactly the recommended dose and got the same result, which is not surprising if what is happening is that aluminum is dissolving until it reaches its solubility limit: the amount of excess solid there doesn't impact the amount that is reached at saturation. I published this new data in an online discussion with Seachem. The owner accepted the results at the time.
Then Seachem tried to prove I was wrong by using their own test. They found nothing, but when I really pressed them, it turned out they used a method that couldn't detect low enough to see what I saw, so obviously the detected none, and still claim that today.
Now with Triton, lots of people are seeing the aluminum and some are making the connection to the aluminum oxide phosphate binders.