2. Bing is one of very very many sources, which are all remixed and combined, so the end result is completely different from Bing or any other upstream source (and most often, far better).
3. No, no, no. Nothing at all is sent to bing other than a plain request for an ad coming from DuckDuckGo servers with the query (for a page-relevant ad). So, this sentence: "user agent (which I can assure you, it does get sent to Bing)" <-- Utterly untrue.
Now, when you click on an ad, naturally your information is sent to both bing and whatever site the ad was for. That's to be expected, you are leaving DuckDuckGo at that point. This is also true when you click standard result links, of course.
Now that that's out of the way, my personal opinion of result tailoring/bubbling is: it's good sometimes. I love being in my cozy little bubble when searching for coding things, and even local businesses and such. I do NOT love bubbling when I am doing academic research. Why would I want to see content I already know and like? That invalidates the point of research, and even gives me a false idea about whatever I am researching.
Bubbling and tracking, while good in some cases, should be optional in my opinion.
Btw, Google does personalize searches in a useful way. You may not have noticed it, especially if you're from the US. DuckDuckGo is unusable for me whenever I search for local stuff. Also, when I search for "Ruby", the first result is about Ruby, the programming language. Guess what my wife sees when she's doing the same search.
You can place "personalize" in quotes all you want, but they are doing it because users want it, period.
Also, I see people here being so enthused about DuckDuckGo, ignoring that the project itself uses Bing's APIs as its primary engine. Of course it does, who would be insane to recreate Google's infrastructure and algorithms when you could get it by piggybacking on Google's main competitor and without which DuckDuckGo wouldn't have been possible.
And here's what a lot of people miss - you don't even need a unique cookie to track users down. All you need are IPs and some smart algorithms for disambiguations. You can't reliably identify all users, but you can identify many of them only by keeping the history of searches per IP. And if you throw in a user agent (which I can assure you, it does get sent to Bing), the reliability increases even further. Throw in IPv6 and we won't be needing cookies at all.
If anything, fear of technology doesn't do any good, because it will happen, regardless if you want it or not. Privacy concerns need to be solved by laws. If governments disapprove, then we need new governments.