In general, I think that if players play something it can't be that bad. Thus, to make a good game one should try to make players play it a lot. In my opinion a grind can be a good grind if the player accepts it. However, not every tool that increases /played is a good one. Loss aversion is not.
When World of Warcraft introduced daily dungeon quests in TBC I welcomed them. I liked that running dungeons would gain me some advantage, because I thought that too few players were running them. I soon changed my opinion when I realized that I forced myself through one dungeon every day.
When Blizzard introduced daily quests I understood their desire to use some techniques of the fabulously successful leveling game in the endgame. And I understood that there needed to be a limit. Daily quests seemed like a useful thing at first. I soon changed my opinion when I realized that I forced myself through daily quests every day.
What really happens with dailies can be explained with loss aversion. Players feel as if they miss a sure gain when they skip a daily (/weekly). By introducing an artificial limit of how often an activity can be conducted per day (/week), Blizzard also introduces a potential permanent loss. If you don't do a daily today, it will be lost forever. Thus, you feel like you really should do the daily.
This is a completely different situation than the typical grind. Since you can always grind as much as you want, you can advance your character at your own pace. What you don't grind today, you can still grind tomorrow. This leads to players playing not as much. Just like you are much less inclined to do your homework today if you can still do it tomorrow.
But it also leads to players who are more probable to play when they actually want to play; and not to avoid losing something.
I am relatively unafraid when it comes to making players grind for something. If they do it they probably like it. Loss aversion, however, is different. It burns players out. Players play even though they hate the content! Now, the moral aspect is one thing. But even for the long-term success of the game, I don't think this makes any sense.
At the end of the day, the company is not interested in players playing. It is interested in players paying. And players who don't play, don't cost a thing. This, of course, assumes a monthly subscription-based business model. Microtransaction-based business models depend on using cognitive fallacies, anyway. Expecting them to not use one as prominent as loss aversion is fatuous.
But a monthly sub model is perfectly fine, if players play only when they feel like it. From the company's point of view, the important thing is not that they play a lot, but that they remember a fun time from the last session and always feel like there's still something worthwhile they could do in the next one. Some goal to strive for. As long as they feel this way, they won't cancel the sub; irrespective of how much they actually play. In fact, the less they play the longer they need to consume all the content (hint).
Thus, to use dailies the way today's WoW does, doesn't make a lot of sense to me. Sure, it makes players play more. But why should Blizzard care? Players don't become more satisfied customers by doing dailies! They rather burn out and stop playing as an act of liberation. These players are much less likely to renew subs or return after they canceled them.
Edit: Sunken cost fallacy didn't exactly fit. I replaced it with "loss aversion".