None of the presumed negative outcomes should be dismissed.
But surely if the proposal is for a trial run of an opt in/opt put system that is to the benefit of the people who see flaws in the "like" widgit.
Becuase the trial will prove them right, MNHQ will knock the whole thing on the head at the end of the trial,period, and in the future when posters ask for a like button evidence can be provided as to why it is not a good idea on this site.
Instead of having to debate the merits of "trolly, MNQueenie accusations, big offence and upset WILL be caused people will be able to respond to future requests for post like/vote buttons with "trolly, MNQUeenie accusations, big offence and upset WAS caused on this site when it was trialed, ergo not a good fit with the sort of people who want to post here"
I personally could hapily live without a like button, I prefer to respond with a post to anything that really tickles my fancy.
I'm here for a chat, advice and a tussle, clicking a button does not scratch that itch.
But if a huge number of posters/lurkers who help keep the site going with their page views would be more inclined to keep reading even greater quantaties of pages thanks to an opt in/out widget, then fine, try it out... and dump it if it doesn't work or it goes down so badly that the people producing most of the content that generates the page views take umbrage and stop posting. Keep it if the worst of the predictions do not come to pass and it improves the attractivness of the site to lurkers so they keep on coming back and consuming pages/clicking on ads.
Although I still think any like button should definatly be anon and have "added value, restricted troll appeal" by limiting the number of posts a registered account can "like" in a week/month whatever.
I think taking it to the next level with tweaks rather than a stright importing of a basic function is more interesting. If you are not a tech. Just the person with the idea, living happily in a code free zone.