When is a "good" time to give bad news?

So The Queen made her big announcement Friday afternoon, approximately 3 1/2 months before she leaves.  Graduate students who are moving with our advisor theoretically don't have to go at exactly the same time she does (although that would certainly make the most sense for most situations).  Theoretically and practically speaking, people probably have about 2 months to decide what they're doing.  While there isn't a particular hurry for grad students to decide if they're staying or going, the longer they take to decide the more time is "wasted" towards getting their degrees.

Some people think she should have told students sooner.  But I'm not sure there really is a good, or even better time to give news that is inevitably going to flip people's lives upside down.  Does having four, or six months, rather than two make that sort of decision any easier?  I'm not remotely convinced that it does, though at the same time, all the reasons for not telling people, especially students, sooner mostly sound pretty selfish.

Obviously for students who are going to switch groups, the sooner they find out, they sooner they can make the change and continue making progress towards their degrees.  Although for students who are going to drop out, delaying the announcement means they have more work done, and have more research done that could go into a masters thesis (assuming the previous couple of months have been "good" lab months).

The sooner students know, students who are going to move are much more likely to make sure they're not going to be in a situation of breaking a lease.  Breaking a lease can be super expensive.  More expensive than most grad students can really afford.

I definitely agree with an advisor not saying anything until the decision is definite.  (Although in this case I'm quite sure that The Queen was sure about this a LONG time ago.)  There's no reason to throw people into upheaval and angst if it may not turn out to be necessary after all.

From an advisor's perspective, it's easy to see why one might put off such an announcement as long as possible.  They're facing a "lame duck" period no matter what.  It's awkward to tell your colleagues and superiors that you're leaving, especially when many people will be upset by such an announcement, and then hang around for any longer than absolutely necessary.  Not to mention you'd only be extending the lab's inevitable down time.  In an ideal world, the lab can be packed up, moved across the country, unpacked, and sensitive/delicate instruments up and running within 2-3 weeks.  In reality, it's probably going to be longer than that.  Productivity will also decline in general well before the lab is packed.  Students who switch groups will do that as soon as possible, and that could mean a substantial percentage of the lab's "workforce."  Anybody who is moving with the lab, but unhappy about it, is unlikely to be productive.  Happy grad student is much more likely to equal productive grad student than an unhappy one is.

I don't know what the "right" time for such an announcement is.  It's all going to be a giant mess no matter what.