2 unrelated ?s for the pros.

Pure Direct or the equivalent will defeat room correction. It turns off ALL EQ and DSP processing. Essentially the AVR takes the signal and amplifies it. I prefer PD for music listening. I believe it takes a sub out of the equation, but I am not sure as I don't run one. If the signal is 2.0, it sends a 2.0 signal. I am sure if I am wrong someone will correct me but I am almost positive. Audyssey is nice to have as it attempts to flatten the response curve of your room. Some swear by it and others think it is a gimmick. YMMV.

I can't get into the balanced vs unbalanced argument too much. But from what I understand the average person wouldn't be able to tell the difference with short cable runs. This falls into the last 1% of sound quality differences that the obsessed will go after. You should worry about Room Treatments and Speakers long before trying to hear the real differences between the two. I believe it has to do with interference in the cable runs. Again, someone will answer with more knowledge if I am far off base.

Depending on how the receiver is set-up, Pure Direct should behave as mbyrnes described (it may be called "Source Direct" or "Direct" or something else entirely though, and there may also be a "Native" or "Standard" or "Auto" mode that will look very similar in literature). In a conventional implementation, the mode will defeat all EQ, tone adjustment, and bass management; it should not defeat channel leveling or time alignment though (if it does, it'll stand out like a sore thumb). The "Standard" mode will get you bass management but defeat correction; both will defeat interpolation (in other words, you can't turn stereo into 5.1).

Regarding balanced vs unbalanced; balanced simply means making use of balanced signal connections which improve noise rejection. For very short runs it likely doesn't "matter" but for long runs or if you're integrating a ton of equipment, it can be essentially (it helps to deal with ground loops, but does not guarantee them solved). It's very common with pro-grade equipment (and if you're integrating pro-grade equipment you should be aware of the level mismatch (I can't find a single good explanation, basically most pro-grade equipment assumes one level (+4 dBu) while most consumer equipment assumes another (-10 dBv - note the u and the v are different) you can get adapters that correct for this; they're fairly cheap). You'll also find balanced outputs on some more hoity-toity gear; it isn't something to run away from, but I wouldn't loose sleep over it. If your equipment supports them, there is no reason to not use them. For example, if you end up with the Marantz AV7005 (which has XLR outputs), if you have amplifiers that will accept XLR in, use them. However if you're using an AVR which only has RCA (single-ended, unbalanced) outputs - it's not the end of the world. And we'll ignore the "truly balanced" vs "XLR output" argument as it usually just gets into "how much money are you going to pay for the same milkshake."

As far as room EQ goes - it will always be available once calibrated when not running in a bypass mode. It will attempt to reach a pre-defined target function (which is not flat in the majority of cases) within the limits of your equipment and acoustic environment; the better your space and speakers are (And their positions) the better job it can do. Also having a more sophisticated measurement routine helps (one that can collect data from multiple points and has precise filters). It will also handle (or attempt to handle) things like time alignment, channel leveling, and bass management (both of which you should verify after the routine completes, often the bass management settings are not fully ideal, and time alignment may be slightly off depending on the measurement).

The "Pure direct" mode for the SR6006 might be labeled differently then, or perhaps it simply forces bass management because of how your speakers are set-up. I'm not very familiar with the unit.

The 6006 is no different from other brands in this respect in that in DIRECT/PURE DIRECT mode there is no bass management, the front mains are treated as LARGE and the only way to get the sub to play 2.0 audio is to set the sub to LFE+MAIN and duplicate the same low frequency audio playing on the front mains to the sub as well (aka "double bass"). Note that it's sister model, the Denon 3312CI also has a separate 2CH Direct setting in that the speakers can be set to SMALL with sub set to YES to avoid the double bass issue.

Pure Direct is different than Direct in that it also turns off the display signals in an attempt to surpress extraneous "noise". My Marantz does not turn off the sub in either Direct or Pure Direct modes. It does turn off rear surround or front height signals (but leaves the surround if it's present).

I meant to add that my system also does not move the fronts to large. So I'd refer to the manual for all of these.

Maybe these manufacturers should explain some of this stuff better, and create a pure direct+subwoofer (or something like that) option.

I think a lot of people new to receivers may not intuitively understand that bass management usually gets turned off in direct mode.

I gave up on such modes years ago UNLESS I specifically want to run my L/R speakers full range. And everyime I try that, I prefer the bass management option. Which is why I think (unless maybe you spend a lot of money on really good full range speakers?) playing music with the subwoofers on and bass management active will work better.

Which means buying full range speakers may be a waste of money? I am considering going back to all bookshelf speakers next speaker switch I do. I gave gone back and forth, and now feel the main benefit to full range speakers is (a) no stands (I dislike stands) and (b) no second guessing whether your bookshelf speakers are going low enough.

(I expect disagreement on my points which is fine, maybe someone can explain where I am wrong.)

I've always heard this argued as a room size thing - basically if you have a bookshelf that can reach something like 60hz in a test bed that doesn't mean it'll do it across a big huge space where a tower might. Then again, I have seen (and own) towers that go over a cliff at 60hz as well, so I don't know how apt I am to believe that rule of thumb (mostly based on experience and stuff I've read).

I think it overall depends on the room, the speakers you pick, as well as the supporting equipment (subwoofer(s), processor(s), room treatment(s), etc). You can get good results with either, and either has it's shortcomings. I like tower speakers, mostly for the reasons you've listed - I also tend to preference stereo listening modes at lower volumes (in a lot of cases the subwoofers won't even trigger - I know, heresy!). Aside from that (admittedly very weird) scenario, I doubt there's a clear-cut (as in non-preferential) answer that fits all users.

Regarding the "pure direct" question - I sometimes rely on the feature (again, stereo listening modes), if I want the subwoofers off. And yes, I really do mean that.