Should Wiktionary avoid use of straw polls?

Some Wiktionaries including the English Wiktionary use polls to gauge editor opinion. In the English Wiktionary, they take place in Beer parlour and use icons to label supporting and opposing votes; rarely, a poll takes place in the formal vote subspace. Should Wiktionaries abandon this practice, in keeping with Meta's essay Polls are evil, and its admonition "Don't vote on everything, and if you can help it, don't vote on anything".

Limitation: Only two of the arguments from Meta are presented below. The debate can be expanded to cover all of them.

Arguments for

 * As per Polls are evil: "Polling discourages consensus. Having the option of settling a dispute by taking a poll, instead of the careful consideration, dissection and eventual synthesis of each side's arguments, actually undermines the progress in dispute resolution that Wiki has allowed. Wikipedia is not a democracy. This is a strength, not a failing. Dialectics is one of the most important things that make Wiki special, and while taking a poll is very often a lot easier than helping each other find a mutually agreeable position, it's almost never better. [...]"
 * That sounds somewhat plausible in the abstract, but does not explain how, say, the disagreement about the use of "from" vs. "<" in etymologies can be resolved using the method of "careful consideration, dissection and eventual synthesis of each side's arguments".
 * The "from" vs. "<" poll from the English Wiktionary is an example of collecting poorly argued votes on the matter: some have empty rationale and some have presented flawed arguments. For instance, multiple votes indicate Wiktionary is not paper although the supporters of "<" did not say anything of the sort but rather pointed to eye scanning. In the poll, there are almost no responses to what people say, so it is not "dialectic"; rather, it is as if a chamber where participants stick stuffing into their ears to perfectly ignore what others are saying. Were this a request for comment administered in a Wikipedia way, this would not have happened.
 * Even if this is granted, the matter is probably of such little weight that making some decision, any decision, enables uniformity of information presentation (professionalism) while causing no harm. And it is unlikely that the poll participants voted against their own preferences; the comments can then be seen as a mere formality.
 * Even if the matter is of little weight, it is valuable to foster culture of trying to be objective, articulate and trying to find objective arguments.
 * Is "eventual synthesis of each side's arguments" even English? What does it mean? If arguments of different sides exist, how can they be synthesized? What could be done is synthesize new arguments from elements of old arguments.
 * As per Polls are evil: "Polling encourages false dichotomy. Very rarely are there only two potential positions on an issue. Simplifying a complex issue to a yes/no vote creates a false dichotomy. For example, in a vote for deletion, the option of merging the article with a similar piece is often ignored. To help counteract this, if you see a third option or compromise that has not been discussed, mention it!"
 * When there are two dominant practices, the use of "from" and "<" in etymologies, there is de facto no false dichotomy.
 * Polling does not need to lead to false dichotomy, presenting only two options. One can present, say, 4 options and use some for of preference voting to aggregate the result.
 * During the polling, the poll participants can indicate other options that were not presented yet.

Arguments against

 * For issues that are a matter of taste, polls are an excellent way to resolve otherwise irresoluble disagreements. For matters of taste, it is often hard to reduce what are to some extent subjective preferences to objective arguments. An example of matter of taste is "from" vs. "<" in etymologies.
 * Perhaps one could in fact reduce matters to objective arguments, if one tried hard enough. One would determine matters not based on "I like it" and "I dislike it" bur rather based on, say, usability testing. Thus, "I like it" would be replaced to "users are likely to like it".
 * The projects do not have the resources to do usability testing on all the matters that need a decision. Thus, as an ersatz, editors are taken as a sample of the population, and the majority of the sample can be taken as representing the majority of the reader and editor population. This is imperfect, but practicable and not too bad.
 * But this can harm minorities, e.g. when a change in information presentation would create a mild burden to majority but help a minority that requires e.g. the user interface to be accessible for people with disabilities. Such a change would be desirable yet rejected in a poll dominated by non-disabled people.
 * Even the alleged Wikipedia argument-based request-for-comment process de facto uses an element of polling. If it were not so, it would not contain boldface "support" and "oppose" and editors would not be adding their votes after a vote that already made their argument. That is not conclusive yet suggestive since the debate is about what should and should not be done rather than what Wikipedia actually does.
 * There is quite a difference between Wikipedia's requests for comment and Wiktionary's polls often containing comment-free votes, showing that Wikipedia's emphasis on strength of arguments makes a real difference in participant behavior.