Maximise your comparative advantage

A snarky mathematician [Stanislaw Ulam] once challenged the great Paul Samuelson to name an economic proposition that is true but not obvious. Samuelson’s choice was comparative advantage, which shows, among other things, that there are mutual gains from trade even if one nation is better than another at producing everything.  

 Here’s a homespun illustration. Suppose a surgeon is also a whiz at house painting—better than most professional painters. Should she therefore take time off from her medical practice to paint her own house? Certainly not. For while she may have a slight edge over most painters when it comes to painting walls, she has an enormous edge when it comes to performing surgery. Surgery is her comparative advantage, so she should specialize in it and let some others, who don’t know their way around an operating room, specialize in painting—their comparative advantage. That way, the whole economy becomes more efficient.  

 The same principle applies to nations. Even if China could manufacture everything more efficiently than the U.S. can (which it can’t), it would still make sense for the U.S. to specialize in the goods in which it has comparative advantages, and then trade with China for the things it wants but doesn’t produce. Both countries wind up getting more for less.

Alan Blinder
A Brief Introduction to Trade Economics 
Why deficits are normal, especially for a country like the U.S., and what is comparative advantage.

It is worth considering the relevance of the concept of comparative advantage to doing science and its administration.

Individual scientists, should identify their comparative advantage [e.g. a particular technique, a sub-field, approach, instrument, ...] and maximise it. Furthermore, they may be better than their grad students at doing certain things, but that does not mean they should do those things.
Some exceptional scientists may be good at public outreach and chairing committees but they need to be careful such things don't stop them from doing the science for which they have a clear comparative advantage.

Collaborations should also make sure individuals maximise their comparative advantage. A collaboration may also have a particular comparative collective advantage [e.g. combined theory and experiment, chemists and physicists, ...].

Yet, there is significant sociological pressure against maximising comparative advantage for the good of science as a whole. For example, the pressure to publish more papers may lead to a gifted individual focusing on low-lying fruit and publons, rather than maximising their competitive advantage to tackle and solve difficult problems with patience and creativity.
The pressure to work on fashionable topics (e.g. to boost citation indices or get funding) may pressure scientists to leave fields in which they have comparative advantage.

Administrators should think twice before increasing the pressure on scientists to do more of their own administration, lab maintenance, public outreach, IT support, secretarial work, ...
These are not things in which they have a comparative advantage.