This rule applies to many activities related to software development (i.e. to architectural design or class design). Possibly, it also applies to activities outside of the field of software development.
Many projects fail to deliver on time, because people are afraid to make decisions when they, in actual fact, have to be taken. People often fear making decisions and stall the process of making decisions. Why? “Well, a decision can be final. I have to bear the consequences, don't I?” Due to this fear no decisions are being made at all. Or sometimes made by the ‘wrong guy.’ Just somebody high up in the chain of command, who has to make a decision -- that is his job. However, more than once he is simply not capable of making the right decision.
Therefore, we mention it again, it is often better to just make a decision –- often time boxing will force you to make a decision -- rather than stalling the process. If at any given time it turned out that there is a better solution, then claim you found it, and validate this decision.