I'm going to answer this without Googling or citing outside sources, because you've already used them (FS Handbook and graphs).. kinda like we were sitting and talking.

The importance of knowing density altitude, is obvliously to account for its effect on performance. In a piston powered airplane, this is triple-important because it lowers performance three ways; Less air for the wing to generate lift, less air for the prop to generate thrust, less oxygen for the engine to generate horsepower.

Everything is referenced from a standard (and this is from memory): Sea-level; 15C; 29.92.

1) Atmospheric pressure readings are all referenced from sea-level. Say you takeoff from a sea-level field and fly into the mountains where there's an airport at 5000msl. The barometer at that field is already calculated to account for the field elevation. The altimeter setting you'd get from their ATIS, is not the barometric pressure AT that field's 5000 foot elevation. It's the barometer reading you get in a 5000' deep valley, next to the airport. If that weren't the case you'd have no way for transient aircraft to be on the same page, altimeter-wise.

Without real-time data, you have to just work from the standard reference. If the field's elevation is 5000'.. then for the planning purposes you asking about, it's density altitude is also 5,000 feet. No need to try to figure out what the barometric pressure WOULD be AT the field.. that's a big "who cares" for aviation purposes.


2) This question assumes that you DO have access to real-time data. Temperatures aren't relative (like barometric pressures).

Whether you're at 8,000msl tryng to calculate TAS (by the goofy dial on the ASI that has you line pressure-altitude up with OAT), or you're on the ground trying to decide if the density altitude will allow you to clear some trees ... the temperature you're concernd about is the actuall temperature right where you are (OAT  outside air temperature).

My pleasure.. truly. I really enjoy seeing people take to simming from a realistic angle... it adds infinitely to not only the fun, but really does help you develop the pilot's frame of mind... and makes simming a never-ending challenge.

I don't care how often you do it, or how proficient you become... A good IFR flight in a simming C172 is a mental and physical workout. Fly real weather, or better yet, set yourself a few cloud layers to contend with, and even a 100nm flight will require you really apply yourself, from start to finish. The first time you spend an hour navigating by instrument alone (keep that pesky GPS off), and realize as you near the destination that your pre-flight work was a litle sub-par.. it will make your pre-flight ritual THAT much more a fun and important part of simming.

Typical scenario for me would be the next leg in a cross-continent journey. I'll pull up all the weather data for the airport I'm at.. and for all airports along that 100nm leg. On a note-pad, I'll record all the airport data (com & AWOS/ASOS/ATIS frequencies), likely runways (per forecast surface winds), pattern types and altitudes, which airports have fuel available.. etc.  Then in a seperate column, all the nav-aids, ILSs etc... VOR radial intersections for the airports... And if weather forcasts warrant it; I'll print out any approach plate I might need. (no cheating allowed.. once you start the engine and taxi.. you cannot go back online.. you had better have everything you need, in your cockpit).. 

I'll always get an ADDS snapshot of winds aloft, too.. looking for anything that would have a significant effect on my headings and ground speeds. This is rarely an issue as I almost never takeoff with minimum fuel.

Running real-time weather throws enough surprises at you, to keep it interesting. You might full well be planning a VFR approach, but as you descend and visibility fades.. you sure are glad you have those approach plates..lol

For extra fun, if I have time.. I'll Google about the town where I end up landing, and do a little virtual sight-seeing