Guess you've never been involved in software development, right? Fixing one bug can introduce other bugs...bugs not seen before nor expected. That's what happened here. And, it wasn't just Lightroom that got hit with this bug, it also was introduced in Camera Raw 8.1. So, it wasn't really a LR bug it was a LR/ACR raw processing pipeline bug that got introduced in the last several builds that sadly, none of the internal testers caught. Was it a bad bug? You bet...was it "incompetence", to a degree, yes...some code was changed that introduced this new bug and nobody caught it. Was it "disregard for their customers"? You gotta be shyting me...do you honestly think so little of Adobe that this possibility is real in your mind? Or, are you simply trying to spread fear, uncertainty and doubt (FUD).
Look, it's ok to talk about what could have been done in LR5 or what should have been done in LR5 and what may be done in the future (LR6), but pulling this kind of crap out of thin air helps nobody...
I'm sure the engineers that modified the code and caused this bug are busy trying to fix it...is it a "data loss" bug? Nope...but yes, it's a problem but one that can be worked around...is Adobe embarrassed by this bug? You bet...nobody in their right mind however, would say that this bug would be a result of "disregard for their customers".
So, the work around isn't "convenient" but is it a reason not to use LR5? Nope...it simply modifies the way you need to use it. And it impacts not only LR5 but ACR 8.1 as well. So, until the bug is fixed, don't downsample images to less that 1/3 of the original capture size in Export...
In retrospect, I was responding to someone else's information, and I don't know how reliable that information is. I should have qualified my comment to say what my opinion would be if the other person's statement was well founded. Your comments seem to indicate that the information does have a reasonable basis in fact.
No, I don't know anything about software development. And why should I? I am paying for someone else to do that work and would hope that they would do their job well and deliver a product that is free from major bugs for the majority of users. I realize that the developers can't possibly test for every system on which their software might be used, but isn't that what beta testing supposed to help with? If changes are made to the software before release, shouldn't that warrant another round of beta testing, and, assuming the output sharpening bug is real and widespread, shouldn't such a fundamental feature be caught in a round of beta testing?
As for a workaround, the only one I know of is to do the output sharpening in some other software or an earlier version of LR. Your statement that those who are experiencing the problem should avoid downsizing to 1/3 or smaller leaves me bewildered, since a great deal of the output (certainly mine) is for the Web, which typically requires downsizing to that degree.
Simply put, if that bug would affect my system, I would not want to use that software at all, since that kind of output size is something I do frequently and LR4 is working fine for me (now that I have a new computer).
I hardly expect any software to be completely immune from bugs when it is released, but it seems reasonable to me to expect the software to not have bugs that affect the most widely used features and settings on a significant majority of computers, especially brand new ones.
I have gotten the distinct impression from my own experience and from the comments of many other users of LR that Adobe really could either do more beta testing of LR before final release or should be more conservative about adding features so that a very lengthy period of beta testing will be less likely.