Maybe you refer to me as recently I posted something very similar to your quote. I do not like the idea and I said it is confusing but can be easily modeled in Spider Project, still for some reason some people are still hooked on this concept. To me it only make sense on pure duration activities such as concrete curing. If the activity requires 7 days for completion then after 3.5 days it will be 50%. The reality is that these type of activities are not the norm but the exception.

There are other methods to estimate percent complete, one is using man-hours, another is using volume of work and another is using cost. Perhaps volume of work makes more sense but what when the units for volume of work of your activities are different, then how you average percent complete? Even if the volume units are equal maybe the effort to get a unit completed differs.

For the above stated reasons I do not like to use % completes, and the use of a single % complete type I consider even worse. To understand the progress of your job there is no such thing as a magical % complete type or a single magical metrics. Your report shall include a combination of these values along with some summary schedule showing critical and near critical activities. It shall include a narrative explaining what issues are relevant, some visible others not visible.

Any of the available metrics is better than using duration % complete, you are right.