I thought I might paste my reply from technical as it applies here too:
Would one believe that the sensors in MAFs could be similar and interchangable?
The answer to all such analysis is to monitor the adaptation values. The P017x codes are all for exceeding the adaptive capability of the two separate adaptation systems; partial load and CTP (closed throttle). By watching the values the type of problem can be surmized and the solution proven. A code that says the CTP adaptation is at the limit for correcting a lean condition (additive value at 1ms or above) is likely to be a vacuum leak.
A part load value of a multiplicative correction of 1.32 is usually a MAF. Substitution of any piece and a short drive after reseting adaptation to zero will verify the solution as adaptation of ME systems happens in a few miles.
For those with real diagnostic capability one can do it scientifically. Go to a search engine and find a Volumetric efficiency calculator (they were first devised for deciding which size Holley double pumper to put on your small block). Run the engine full throttle (under load) while monitoring MAF, rpm, and intake air temp (this is best accomplished on a dyno or with a graphing scanner see:
http://www.autoenginuity.com/ ) The calculator will need rpm, maf (in proper units), cid (in proper units) and intake air temp (in proper units). The answer should be above 70% for most engines and above 80% for most 4 valve engines and as high as 100% or more for turbo or supercharged engines.
As to the question about adaptive codes and catalysts. An interesting thing happens with a restricted cat. The MAF meter reading is an average of the intake of all cylinders. If one side of the motor is restricted, those three or four cylinders are using less air. Since the MAF only reads total it has an averaging effect. The amount of fuel that is calculated is too small for the free breathing side and too much for the restrcted side. The readings for adaptation will then be plus for one side and minus for the other. If the MAF is already weak then the lean side (the free flowing side) will blow the limit. In this case a MAF will fix the problem for a period but adaptations will continue to be significantly different side to side.
One other bit of info in this thread that is wrong for MBs is the total multiplicative adaptation. Twenty five % is common with many cars but MB goes to 32% on ME cars. That is probably due to all the problems they had with 18% in early LH and HFM cars.