5 Plus - GPS Distance coming up as much a 20% short in the woods.

Guys, actually this issue isn't F5 (or F3 or any other GPS gadget) issue.
Let's try to explain how distance (and mean speed also) is calculated, for example, with 1-second polling rate:
1. Get coordinates of the current point.
2. Get coords of next point (after 1 second).
3. Calc the vector and add it's length to accumulated distance.
4. Repeat above steps again and again.
It's easy, but!
We must keep in mind, that acquiring coordinates, especially in complicated conditions (city streets or forest is classical example), is statistically not guaranteed for every second.
Moreover, because of miscalculation, caused by the GPS technology itself, such subsequently acquired vectors can be randomly spread +/- 10 meters even if you moving straight along the line.
To avoid erroneous point calculation, all GPS receivers implements a kind of predictor inside.
In the wrist devices it works like simple 2-points extrapolator (which is caused by limiting of energy and so calculating power).

And now, lets pay attention to the distance / time boundaries to our model.
If you are runing a trial in the forest, you can move with speed of 3-5 meters/sec. As for me, it's usually 16-17 kph (4.5 meters/sec).
And in the forest your movement trajectory definitely will significantly curved!
So, for one second you can move ahead by 3 meters and move right by 1.5, and next second move ahead by 3 meters and move left by 1.5. Actually it gives 10 meters. But GPS most likely shows you just 6 meters, ignoring you side movements.

I did one simple test - uploaded the GPS track from my F3 into SASgis with detailed satellite view of site, where I can see any curve in my trial path.
There was no surprises. Track cuts off almost all the noticeable bends of the path. In resulting when I calculate the path length manually (by the tool SASgis provides), I got 23% of distance difference!
Of course, GPS one was shorter.

So, conclusion - it's not bug of the specific devices, it's an issue of technology itself.
And there is no wrist GPS (doesn't matter - Garmin, Polar, Suunto etc) which works without this issue.

But as Garmin users, we have good option to correct this.
All you need - the running shoes accelerometer (Foot Pod micro). Wear it on, calibrate it on the precise distance (1 or two miles), and you will get 1% accuracy either for distance and for speed.
It perfectly measured instant speed (far better than non-Garmin devices, btw).
In the settings of my F3 ("Sensors" tab) I set the speed and distance - "always from foot pod sensor".

Time to time I check the the distance and recalibrate the Foot Pod. For new shoes (I mean new model) you need to measure new calibration factor, but it's easy to do.
The distance measured is very precise. Of course, using one more accessory could be annoying, but Foot Pod is light and small. Salomon running shoes have small pocket where you can put it without the frame. And you can forget it for the year or so, untill the CR2032 battery still alive.

I'm running hard trials (across the mountins, with the 350-700 meters elevation) 4-5 times a week, more than 2500 km per year, and the battery lasts for 11-12 month.

Hope this can help you too.