Good points Tom, only a couple of things I would add. I've successfully used the FiddleStick
principle (a piece of roughed up PVC), and also have used the Omnisling method. So, not defending or favoring one over the other. They both have the advantage of minimizing rope grooves. They both require special attention to minimize risk.
The Omnisling can be extended for larger objects with webbing and a rapid. Although you could forgo the rapid, I suppose, the potential of the rope getting stuck in the mechanism may be greater. Not critical, since when you pull/release, the webbing/rapid side comes straight down and does
not go around the anchor. Only the Omnisling goes around the anchor. And if hazard strikes and it got stuck, you can still pull your rope through while the Omnisling remains behind.
Here is one example of rigging an extension:
Attachment 65006
You also don't
typically use a pull cord with the Omnisling, you use the same rope. However, you need half the rope with the FiddleStick (the other half is a potentially smaller diameter pull cord).
Interesting about the softening with use, release, and small rope caveats. I haven't observed them, but I read them with care and thanks. I have used 8mm rope without any problem. I have been sure to make the loop large. Many use two loops. But more importantly to the point, I trust in their relevance so best to take them seriously.
I look forward to further learning all the points of caution for the FiddleStick, some of which may be captured in this thread.