Precision engineering.

Here's an experimentalist complaint for which I do not think there is an analogous theorist problem.  In my lab we have a piece of equipment of European manufacture that is very good and beautifully engineered.  The one problem is, it's so precisely made that it's impossible to service.  For example, after years of repeated thermal cycling, an electrical connector has failed and needs to be replaced.  The problem is, the way the system was put together, there is essentially no slack in the relevant cabling.  They strung the cable through during the original assembly, cut it precisely to length, and then attached connectors that make it topologically impossible to take apart without their removal.  One can't replace the connector without either cutting cabling and inserting more connections, or other approaches with similar levels of inconvenience.  This is the lab equivalent of having to remove half of the guts of a car in order to get to the oil pan.  Ahh well.  Let this be a lesson to mechanical designers:  It's never a bad idea to design a complex system with the possibility that it may need to be taken apart nondestructively someday.