Cross-sectional Schmosh-sectional. What is the issue?

I have stopped accepting "cross-sectional design" as a weakness in descriptions of study limitations.

This is like saying a US sample in 2017 is a limitation because it is not a sampling of the totality of human experience.

My most recent critique says this:

They confess cross-sectional design as weakness, but that actually is not right. 

The cross-sectional design is not the impediment to inference.

The impediment is failure to assess/know temporal sequencing of construct values.

The work might have been constructed with cross-sectional design, to sort out the temporal sequences, thereby avoiding sample attrition, measurement reactivity, and other biases that plague prospective and longitudinal studies.

Framed in this way, the problem is not one of cross-sectionality in design.

Rather, it is failure to sort out temporal sequences.

True, you can sort out temporal sequences more readily in a longitudinal design, but suppose you don't assess temporal sequencing.

All you know is that X converted from a value of 0 to a value of 1, during the same interval within which Y converted to a value of 0 to a value of 1. You have done a longitudinal study, but you haven't kept track of whether X changed from 0 to 1 before Y changed from 0 to 1.

 Clearly, the limitation is not the cross-sectional nature of the design.

Rather, it is failure to keep track of and to manage the temporal sequencing.

Does this distinction matter?

Textbook learning teaches you to think of the cross-sectional design as always inferior to the longitudinal design.

Experience teaches you that the cross-section, per se, is not necessarily a weakness.