While the latter has a vast test bed of real world utilization that reveals PL efficacy, the former can only be evaluated from the standpoint of theory, with a heavy heap of ideology.
I can agree with the broad point that inertia is a major cause for constitutions to stay the same over time. But I disagree with the original commenter's implication that more recently-written constitutions tend to be better than older ones due to lessons from the past, because of the idea (perhaps relating to yours) that contemporary political and ideological considerations can introduce weaknesses into more recently-written constitutions, which overlook potential lessons from the past.
https://en.wikipedia.org/wiki/Comparative_politics
Lest you think this is ideological or theoretical, no -- it's entirely about collecting data of real-world political operation, how similar or different constitutional structures result in similar or different outcomes given similar or different populations and histories.
It's incomparable to PLs, which have perhaps billions of case studies on which comparative analysis can be done.
Because there are probably only ~200 PL's with truly significant usage, and they tend to have a variety of purposes and be used in historically different contexts -- much like constitutions.
The same arguments that programmers have over declarative vs imperative is much like the arguments political scientists have over parliamentary vs presidential. Static vs dynamic type checking is much like civil law vs common law. Perhaps the rise of congressionally-created federal agencies is akin to the rise of object-oriented programming?
And there's similarly nothing "scientific" in proving that one programming language is "better" than another, but people have strong opinions, and experience will tell you which ones are good ideas to pick for a project and which ones are not.
By the way, I'm not sure you know what a case study is -- it's a report written by a person. Nobody's written billions of those. And you don't do comparative analysis on case studies -- a comparative analysis is a case study, where the cases are countries (or programming languages).
