My proposal is, just read both reports. The old one dates back to 1999. In the text of report 71 you will find hints to differences, for example measuring depth in water etc.  It´s easy to download the old version. I append it for convinience.

@Hanno, Report 67 is the TG51 protocol for dose calibration, while the TG71 report is on monitor unit calculations.

@Intan, there was no prior AAPM report dealing with MU calculation. If you follow along with the report, you'll find that it recommends an approach that is very similar to the familiar Khan approach (from his textbooks). I'm not sure I really understand what your questions are referring to, so you'll need to be a little more specific.

usually for SSD technique, we used PDD at zmax. And then, field size determinations for output factor (Sc,Sp)  and for PDD in SSD technique is equal. However, in TG71,there are so different. TG71 recommended for used PDDn at depth 10cm,

I thought I had answered this a few days ago, but I guess not. The key is to look at the derivation figures. The calibration is at a depth of 10 cm (at least for the AAPM protocol), and things like in-air output (TG74) and in-phantom output are measured at 10-cm depth (or should be). If you want to use dmax for your MU calculations, then you need to have output factors and calibration at dmax as well. It's not wrong to do that, but you have to first make the choice of depth and then try to make everything consistent after that. If you wanted to have all of the output factors at 10-cm depth and still use PDD(dmax), then you will find the you need an extra PDD(10) in the equation, and then the equation doesn't look as similar to the TPR equation as it could be. That's why PDDN is used.

AAPM TG 71 can be applicable for cobalt-60 for treatment time calculation but it wholly used for LINAC for MU calculations. in report MU can be replaced by Treatment Time in formalism.