Due to the nature of human existence, people miss posts when they first appear, and people find valuable posts through the search function at a later date. Time should not place a limit on the wisdom a thread contains. That's one of the advantages of the internet: make a post, come back a month later, and it's got a lot of wisdom added underneath it. That's especially true of obscure subjects, and even truer for "unsolved problems." (The kind of post that most needs wisdom, whenever it arrives.)
When someone perfectly and concisely states a problem, their post should slowly accumulate posts over a long time period, and eventually, if their post is optimal, it should contain nearly all the wisdom the system can muster. But this isn't the case when comments are closed. This inherently causes "wasted motion" in the system, because people must repost the same problem many times, even when the "most expert person" has already well-covered the problem. (And even in cases where there is no problem, there is relevant wisdom related to the post that shapes and enhances it.)
Why refine and edit something to high standards if you're going to post it in the middle of the night, and commenters are going to be shut out of commenting on it in a few short days, weeks, months? I've posted on long-term posts on the internet years after they were posted. This incentivization also diminishes the incentives placed on optimizing the entire steemit system. After all, if the interactivity of the thing is artificially limited, why value its creation, and why refine the standards by which future posts are created? (For example, I view the "five tags" limit to also be incredibly short-sighted and, well, Kurzweil's functional definition of stupidity (unwittingly self-destructive). The limit is stupid, because it's a very-limiting limit.
Let me first say that I understand why keyword limits exist: So that people won't simply type in "all the most popular, attention-getting keywords" on every post. But this can easily be "gotten around" by auto-generating all the keywords, both "recurring often in the article"(used a lot in the post in question) and "unique to the article"(rare in other posts) as well as "likely typos or alternate spellings of keywords"(so the post can still be found if there is an error in the human search). Adding these categories and allowing a user to select which words appear as keywords would help optimize the ability of others to find posts they care about, while doing nothing to diminish the quality of people's ability to filter through posts. Additionally, it would allow an author to select multiple forms of a tag, so that those searching for a rare form of the term won't miss it if they type it in wrong/differently.
I believe that, even if tags or keywords are going to continue to be limited, that the limit should be increased to around 30 or so words, in proportion to those words' use in the post. Additionally, probably 2/3 of the keywords should be autogenerated, and the author allowed to select how many, if any of them, are added as tags. I envision three columns for these auto-generated keywords, as noted above.
The tag limit is limited by a too-narrow vision, and not helpful. With tag limits, come additional tag limits, because nobody wants to optimize a feature that is limited to 5 words. Expand the number of tags possible, and perhaps there are different categories for tags, as well as value ranks for tags, as well as "which single tag is most important?" selectors, etc. Or perhaps those things all go into the same generic "tag basket" ...but we never get there without more than 5 tags possible.
People like Kevin Kelly, K Eric Drexler, and Vernor Vinge noted long ago that the time-insensitive nature of the internet could dramatically expand the power of slow, serial, imprecise human language, ...if we let it. Their vision of the internet has largely not come to pass. The internet remains, well, ...stupid in many ways. (Or, at least, far short of those visionaries' vision for it.) It's transitory, it can't keep a post in the same location for long, it breaks links, pages, and protocols, it doesn't treat human language words like anything more than "words" or "dumb labels," etc. It is treated like a printed magazine, not ongoing cybernetic texts.
When I hover over a complex human word, why doesn't that word have a "hover-over" that explains the leading thinkers on the subject (relative to the context in the post), in the form of several branches, chosen by the author? This exists on some websites, but not others. It's a programming feature that one can add to "all webpages," (only as a dictionary/encyclopedia) but then the author control is lost. This is just one "missing feature" that was assumed would be all over the place in thousands of different forms, all in competition with one another. As it turns out, it's almost nowhere, due to patent trolling and other bad reasons.
...But that's no reason to limit steem's functionality to be less-than-optimal.
