The Cunctator wrote:
Let me make a practical example: say I, a longtime user, want to create a new article but I'm not at my home computer. To do so I have to log in again; I have absolutely no memory of my password. Before Monday, I could still easily create the password. Now I have to have the system send me a new password, open up my email, enter the new password, and remember to change my settings at home with the new password. These are new hurdles that I may not want to go through, so no new article.
Ah, now I see: This "experiment" is, truth be told, an Alzheimer prevention program to make you remember your password! :-)
Seriously: While a scientific experiment is open in its outcome, it is in some way designed along an expectation, namely that your working hypothesis is correct. Only if the experiment shows that this hypothesis is false, you will consider altering or abandoning it.
I think that's the case here as well. While the outcome is open, the expectation is that it will help wikipedia and, thus, stay. If it turns out that, *unexpectedly*, it does harm to wikipedia, I'm sure Jimbo will be the first to get rid of it.
And for the situation you mentioned: Why not create a user account and write, on the user page, "This is User:XYZ who forgot his password"? That way, the new article can still be credited to you (through the user page). If you, however, would create a new article as an IP, it would not be credited to you.
Magnus