Now I understand what's going on with the data error near Asgard Head, I am reflecting on why the peaks that this definition is identifying in the blue mountains (the area I am most familiar with) seem a little disappointing somehow. Apart from Cloudmaker it is excluding all of the 'classic' bushwalks, and many of what it does include are unnamed and/or on a road.
Considering why this is so, the first reason identified is the 1100m cutoff. Much of the the Blue Mountains National Park, and the areas where people live, is a sandstone plateau that has tilted up out of the sea and erosion over millions of years has carved many interesting shapes and peaks. But only a small part of that plateau exceeds 1100m, so that rules out e.g. Mt Solitary is 919m, Mt Colong is 1047.
Then west of the sandstone plateau is granite country, which has not eroded as much as the sandstone. That means 1) it is higher (so above the 1100m mark) and 2) the elevation changes are far more gradual. So gentle rolling hills, not steep escarpments. The gentleness in elevation change means high points are not particularly distinctive (hence many unnamed) and also means there are no impediments to the tendency for paths and roads to follow ridgelines (hence mostly on roads).
The conclusion I have arrived at is, the 'ABEL' definition will not result in a set of points that are a (subjectively) "interesting" challenge, when applied to the Blue Mountains. Could it be tweaked ? Let us consider...
The easiest tweak would be to reduce the ASL cutoff from 1100m. A 1000m feels if anything less arbitrary than 1100m, and would bring in to play Mt Colong and Pantoneys Crown. It still excludes Mt Solitary, and Mt Dingo/Splendour Rock, which doesn't feel right but lets ignore that for the moment. What tells me we have more work to do is the fact that changing the ASL cutoff doesn't help at all with Mt Guouogang. Yes I know it's a god-forsaken labyrinth filled with ti-tree and blow flies, but it's also the spot that appears to be the high point for the area, regardless of there being a contour line that will (if you trudged long enough along featureless ridges and country roads) let you get from there to the mobile tower on Mt Trickett without descending more than 150m.
So for me, any interesting definition for peaks would need to have Mt Guouogang in the result. We could achieve this by shrinking the prominence requirement to 100m. We now have many more 'interesting' peaks, but at the expense of also adding lots of 'uninteresting' peaks. So what makes a peak 'interesting'? I suggest it is a combination of aesthetics (the view) and physical challenge. A walk with a view, basically. Hmm... a place where you walk, with views, sounds a lot like a national park. What if we used that as a filter?
I hereby propose a definition of a HILDER to be: "a point within the Blue Mountains NP or Kanangra Boyd NP that has prominence of at least 100m relative to all other points within those national parks". Note that there is no ASL cutoff.
@awmross - would you be up for tweaking your ABEL code to see what my definition above comes up with? If it helps there's a KMZ file with NP boundaries in it here: http://www.geocachingnsw.asn.au/index.p ... n2010.html