Howard,
It appears you are getting less than me, so I guess you aren't yet fully run in (Mazda seem to define this as first service or 12.5k miles). At 33k miles it would like to thing I'm fully run-in and, to be fair, it has gradually improved. In similar circumstances as previously mentioned, on that trip the true figure was 45.5MPG.
Perhaps worth me explaining why I'm so focussed on this issue. Mine is a company car. Normally, we wouldn't go above 2 litre for a diesel, but I do have pretty much a free choice what I drive. The company pays the fuel so I repay private mileage at an "approved" rate/mile. That rate, set by HMRC is determined by engine size, not MPG. So on anything above 2 litres the personal payback is more per mile - you could pay the same for a CX5 diesel as you would for a VW Touareg V10 diesel or a V8 Q7, which is crazy. However, as the approved rates are recommendations, you can pay the lower rate if the vehicle is more efficient - I used this to justify the choice, especially with both the 150 and 175 BHP versions having the same figures. Keeping the detailed records proves it in case the taxman does an audit.
My real issue is not what it does - frankly, 41MPG overall average for a 2.2L, 175BHP 4WD (sorry AWD, Mazda don't class this as a 4WD...) is quite good. My issue is that Mazda make such outlandish claims in the first place. The problem is that the universal testing system for official MPG/CO2 figures is set up in such a way as to encourage the manufacturers to make these claims without any possibility of comeback from the buyers.
Still considering buying a plug-in chipset, though...