Nope. The Guardian Link buffers data for a max of 10 hours. I base that on the excerpt below which is quoted from a CGM and Sensors FAQ section on the Medtronic Diabetes Australasia web site. Note the bits I italicized.
What happens when the MiniMed® 640G transmitter and insulin pump move out of range?
If the transmitter and insulin pump are too far apart (about 2 metres or 6 feet), an alert will sound and either a “LOST SENSOR SIGNAL” or “SENSOR SIGNAL NOT FOUND” message will appear on the screen. The transmitter holds 10 hours’ worth of memory and if the transmitter and insulin pump are too far apart for longer than 10 hours, reports may have a “data gap” where information is missing during this time period.
I’m not sure what the heck your medtronic rep was up to with that allegedly “7 day cgm trial”. I mean, yes, it was a trial of wearing the do-hickey for 7 days and also of whether the battery is actually going to last 7 days. But the only data they would have gotten would have been the last 10 hours which the transmitter had buffered.
The other aspect of it is that they would have needed to have access to at least one or more meter BG readings during that last 10 hour period to allow them to calibrate and “make sense of” it. Did they also download your BG meter when you came back in 7 days later?
Is there any way you can contact the rep and ask her/him, “Dude??? What the hell was up with that???”
This is the first time I have ever heard of a “trial” of the inserted sensor & transmitter without also carrying the pump. Even if they didn’t want you to use the 640G to pump insulin, you still could have carried one which you could have calibrated and used to monitor your glucose levels. And they could have then downloaded the 640G’s data which would have given them results from the entire 7 day period. 
Why would they do this? What the heck were they trying to achieve?
I am now curious as all heck about this. 
