from: http://www.autonews.com/article/20150929/OEM11/150929807/epa...
So the manufacturer apparently does not get to chose the vehicle, which is one reason why the cheating code was present in all VWs with that engine model rather than just in the vehicles used for the testing.
For examples, see http://www.beuc.eu/great-fuel-consumption-scam#thescam and https://www.youtube.com/watch?v=EKUtiL-R4sU (subtitled in English)
Car manufacturers do not make the measurements themselves, but they directly do pay for it, and do shop around for the best testers. That raises doubts as to the independence of those testing firms.
Are you implying that it's a bad thing that a manufacturer is forced to implement cheats on all cars rather than just provide a rigged car for test?
Like if you were to say "So we're apparently not allowed to bring guns to school".
You can find alternate uses for many words and then you can go and spend many hours debating your reading of what someone wrote but given that I've (1) already answered your question with my response and (2) those do not seem to be the dominant meaning of the term I think that further debate of this particular matter is pointless.
Pointless -> 'not useful', not 'without a sharp pointy end'.
They also have drivers who are above average. That will help with manually shifted cars.
And of course, they pick the best car they can produce, and prepare it optimally. Tyre pressure will be optimized, the car will be perfectly polished, if that helps, etc.
Cheating starts when they tape shut doors to decrease air resistance, or when they explicitly detect "test is being run" and act upon it.
IMO, the real problem is twofold: firstly, the existing tests were designed in such a way that their results do not reflect what happens in the real world, and secondly, some manufacturers went criminally far in optimizing for it.