monty, Mercedes would have known roughly what tyres were being tested (or should have known) through a combination of their telemetry and what was being said before the test about tyres. They knew the 2013 ones could not have a compound change - only a belt change - which means that only certain alterations would have been possible. The 2014 tyres could and almost certainly would involve compound changes because tyre suppliers do that every year, in response to research and the changing needs/wants of their series. Further, it is likely that at some point the current tyre type would have been used for validation purposes (how else would Pirelli know the new belt type was safer than the old one?), which would clearly have behaved the same as in the Barcelona race weekend Mercedes had just done (adjusted by alterations in weather).
It is known that Rosberg drove differently in Monaco to earlier races, particularly in the early phase of the race. Specifically, he drove much slower in the early phases of the race than the team was doing hitherto. When it is more likely he would have discovered this: in 80-100 laps of practise/qualifying with alternate goals and traffic, or 214 laps in a three-day exclusive test specifically to test tyre performance? Practising such a technique would be impossible to ban in such a test, unless every team is going to be obliged to use robots as testers. (This is vaguely related to why in-season test bans hurt drivers - especially developing drivers - more than teams. It's easier for a driver to improve their performance without knowing all relevant parameters than it is to improve a car).
My guess is that your blind tyre test was in a situation where your tyre supplier had rather more freedom to experiment with tyres and was in pursuit of a variety of goals (not just safety and compliance with series needs/wants, but grip, endurance, wear patterns...) Therefore the contextual cues that Mercedes could and should have used were absent for your test. No wonder your team struggled to get any useful data for its own purposes, when the tyre company provided so few clues compared to Pirelli in the last few weeks.