Hmmn, your link doesn't go where you think.  It's going to a line6 video on creating a preset.

http://www.youtube.com/watch?v=jP-AkfWsZnw

The HD500 can certainly be complicated with all the flexibility and options. The HD300/400 are designed for much simpler "out of the box" playing. Many users choose that route - that's why those products exist.

Well, there are some things I think Line 6 did well, and some other things I think they didn't do well.

The input selections I think Line 6 did as well as they could.  The impedance choice gives you the "auto" setting (which I believe is the default).  As far as the Input 1 and Input 2 defaults, Input 1 all inputs and Input 2 same actually makes the most sense as a default - you can plug anything in and it'll work as expected for both single and dual amp tones.  If they defaulted to Input 1 Guitar, then people plugging in a Variax or Mic or Aux input would be like WTF.  This might mean if you just use guitar you're getting additional noise, but it makes sense to prefer extra input noise than for it to simply not work.  Input 2 Same is necessary to route tones to Channel B so dual tones work.

Contrary to what some people say, Input 2 Guitar/Same doesn't add additional noise, it just adds volume to the signal hitting the amp in a single amp tone (or first mono FX block).  The only reason this became an issue is because the cleanest amp in the unit doesn't have as much headroom as people would like with its default settings.  The real issue is the amp model, mainly its defaults, not the input selection.  I'll get to this in a second.

Routing I think they implemented correctly; however, the documentation and display isn't informative enough and can be confusing.  It makes sense to have mono FX, and they'd have to mixdown the stereo signal to work, then split the signal back to equal stereo.  Theoretically they could have treated mono FX as 2 separate mono FX operating in parallel, one on the left side and one on the right side of the signal.  The downside - this uses much more DSP.  Plus it could introduce its own confusion.  I think they should simply have some indication that every line in the unit is a stereo signal, and it shows what happens to the signal for each effect.  For example, just make the lines doubled up like =, then if the effect is mono in, where it's going to mix-down the signal do like =>- going into it, and if the effect is a stereo out, just have two lines coming out of it.  If it outputs mono which then is split, put a -<= coming out of it.  Simple but makes sense.

Here's where I think Line 6 messed up.

I like the mixer, but it's basically makes the amp volume knob redundant.  This is also confusing because it doesn't affect the amp's tone like a real amp's volume knob would. I'd rather just have the Mixer Channel Level controls, then have a single patch volume knob at the VERY END of the signal chain.  All the amps should output the same volume, regardless of where you set the Master Volume DEP.  This lets you mix the relative volume of two amps for a dual amp tone, but if you want to level your patches, you only have to tweak one knob per patch, not dig into the mixer and adjust both sides, or change both amp's amp/channel volume setting.

Also, the default values for some of the amp models are poor.  The Blackface Dbl is downright stupid.  The cleanest amp in the unit and you default the Master Volume DEP, which casual users don't even know how to get to, TO 100%?!?!?!?!!?  SRSLY, WTF?  Moving it down to 20% and Bias X to 80% cleans up the amp a good bit without sacrificing virtually any of the tone.  Given the amount of complaints about this amp's headroom, just change the defaults in the next firmware!!! It wouldn't change the tone for users' patches who already dialed in their patches, only those tweaking new tones.  There are other such examples - Line 6 Elektrik Master Volume at 100%???? It sounds like absolute arse at that setting - it certainly isn't the setting baked into the v1.1 Uber, which it was based off of (cloned more like it).  Also, I feel like the Marshalls should have Bias set lower, but maybe that's more accurate.

The cab/mic simulation I think was implemented a little unwisely.  It sounds like they were created with cabs/mics being pummeled with volume, forcing them to drastically alter frequency response.  I understand you want to do that to get more of the nuances of the gear, but then any engineer is going to have to EQ the resulting tone to get it to sound back to what the desirable frequency response is.  So either model the cab/mic combinations at lower volumes or EQ them a bit after the fact to make them more consistent.  The difference between the SM57 on and off axis is too drastic, as is the difference between the XXL and Treadplate cabs.  This results in having to EQ the tone heavily, and given the EQ options on the HD, you're likely to have to use a bunch of them in a single patch to get it where you want it, but even then the tone ends up sounding a bit washed out at certain frequencies.

Another point of confusion is the output modes and the "live-voiced" cabs (output modes other than Studio/Direct with a cab selected). This shouldn't be called "cab simulation" - it's really more of a cab match EQ.  But in this case you still need to know the actual speakers being used.  IE, if you're using an actual Mesa Recto cab with V30's, selecting Treaplate 4x12 shouldn't change your tone at all.  And for this mode, it should indicate that the mic is disabled instead of having it appear it is still active as you're editing.  So it gets really confusing.

Here's a better way to do it.  Ditch output modes all together.  Separate the amp/cab/mic block into separate amp and cab/mic blocks.  If you want the simplest setup of sending the pod tone to a real amp, just disable the cab/mic block altogether.  Turned on you have two modes - match cab or cab/mic modeling.  Match cab has desired tone and actual cab used selections, which works as described above.  cab/mic modeling requires the selection of both a cab and mic, the combination of which loads the IR used to recreate that cab/mic simulation.