Honestly, that's what I began looking at also. Django and Rails are compared frequently - even if those comparisons aren't hostile.
This is the first time I've seen this benchmark, and it was really really interesting to see how Djangos performance degraded as more queries were executed. They mention that the lack of connection pooling is likely a big factor. I never realised how much that could affect an application.
Once you wrap your head around one (embarrassingly, it took me a while to really understand what celery was doing -- I had an easier time getting comfortable with the multiprocessing library), it shouldn't be difficult to pick up another and compare them locally.
Celery does what it's supposed to; that's about all I can say for it. I'd suggest writing a few simple jobs and testing them in similar environments under as much load as you can. Look for things like consistency of logging and running tasks exactly once across multiple consumers despite the semantics of the queue you pick.
I'm in the process of writing a pet project using Django and would really appreciate your insights on what packages you find useful (as mentioned at the bottom of your post).
It's not a Python package in the sense that you might mean, but using virtualenv is absolutely fundamental to good Python development (at least the kind that lets you keep your sanity).
South is also exceptional; however, the developer recently ran a successful Kickstarter campaign dedicated to creating an improved version intended to be merged with Django itself, so that's something to watch for.
Django-CMS has proven invaluable for many of my projects and clients, though it's very heavy and more of a project-tacked-on-to-your-project than just a package. It's great to know, though, and very powerful.
django-registration can be useful if you're into full-on user profile usage, though many sites don't actually need this. There's a separate project that provides a set of default templates for this package, and it's almost more useful than django-registration itself.
livesettings is an outgrowth of django-cms that's pretty useful when you need what would otherwise be settings.py constants that can be modified at runtime without a deployment and server restart. Be careful, though, as overuse of this means you might have design issues.
feedparser is a very, very easy way to consume XML feeds.
If you want to parse something in a very user-friendly way, BeautifulSoup is hard to beat. If you're into the nitty gritty and/or need serious performance, lxml serves the same purpose but kicks far more ass.
If you need to interface with Amazon's AWS environments (EC2, S3, etc.), boto is a brilliant library.
These are more Python-in-general than Django-specific, but that's often a good thing.
While all suggestion are useful the biggest take away for me is livesettings. TBH I can't think of how I would use it at the moment but I can see that it could come in handy later.
I actually don't "know" either. But I suspect it has to do with knowing which python to use (system generally) and if the user wants to use global packages as well.
Honestly, I have no idea. I haven't ever tried to move a virtualenv, and it's not something that's supposed to be part of your repo (you should build your env separately on every box; it's intended to encapsulate dependencies from the system, not to transfer them around with your code).
If your system gets updated/upgraded to the point where your virtualenv breaks, rebuild it from scratch; all it should take is a bit of time.
[edit: I am not a developer involved with virtualenv, and the statement about what it's intended for probably shouldn't be quite so strong, as this is only my impression of it and the way I use it.]
Having transferred companies to both O365 and GApps from Exachange I really have to say that O365 is a horrible product. It is slow and nearly everything about the back-end seems like it was put together in haste.
