Re: Where's the red line?
The other problem here is that too much credence is applied to the test results.
If you have one car that tests at, say, 120g/kmCO₂ and another that tests ar 150g/km, you could reasonably assume that the second car will be worse in real-world conditions.
If the two figures are 120g/km and 121g/km it's a different situation, the test is likely to be too unrepresentative of the real world for these results to be considered as meaningfully different.
That doesn't stop a government putting the second car into a tax band that will add hundreds of pounds to its purchase price, and possibly hundreds of pounds per year for vehicle tax. As a consequence there is huge pressure on manufacturers to stay within a band, which inevitably encourages this sort of shady behaviour.
Much the same problem is seen with things like stamp duty on houses. A true sliding scale, instead of 'bands' could help reduce this effect