There is nothing wrong with writing tag wikis if someone wants to. That said, I do not find it “unfortunate” that many tags lack wiki.
For some tags, it is very desirable to have some explanation. For example, the meaning of the tag qma is unclear to those who are not familiar with the complexity class QMA, but renaming it to something like “quantum-merlin-arthur” makes the meaning of the tag less clear to those who are familiar with the concept, so using the tag wiki (actually its one-line summary which is shown on the question list page) is very appropriate. Another example is csp; renaming it to “constraint-satisfaction-problem” might have been better, but the system does not support this long tag.
Some tags have closely related tags which may be more appropriate for particular questions (such as approximation-algorithms, to which approximation-hardness is a possible alternative). Other tags deserve some explanation of usage (such as big-list and open-problem). I do not think that tag wiki is a very effective place to write this kind of explanation, but we do not have many choices anyway, so I think that writing tag wiki is one of the best options available.
Other than these cases, it seems to me that tag wiki is completely optional. As I said, there is nothing wrong with writing tag wikis if someone wants to, but I do not think that tag wiki is very much needed.