There is value in discussing publically the safety status of CTL procedures. It may be worthwhile to use the vocabulary and concepts which technical safety analysts generally use.
For example, there is no worthwhile division into "safe" and "unsafe"; as in "CTL manoeuvres are safe"/"CTL manoeuvres are unsafe", for there are no generally-accepted criteria which allow such a binary distinction. The pertinent question is that of risk: how risky is CTL manoeuvring?
Let us grant that, as well argued by many here, CTL is defined by a set of procedures that, if followed to the letter using well-trained, almost-perfect judgement, is extremely unlikely to result in an accident. However, it may be one of the few instrument-flying manoeuvres which is intuitively more risky in large transport aircraft than in Cherokees (I bet the numbers don't bear this out, though).
Specific questions of human reliability /engineering psychology then come into play. How risky is it for
your airline, with
your pilots, not faceless licence-holders but Bill, Andreas, Ahmed and Soo Li, to engage in CTL at airports X, Y, Z? That generally cannot be answered, but the question is well-posed. The question, and the corresponding judgement, is one of
risk.
Some argue here that CTL is not more risky than other manoeuvres, for competent/well-trained pilots, for competent airlines with good training regimes and other caveats which in other contexts are called
ceteris paribus conditions.
OK, but do those ceteris paribus conditions obtain generally? In
The July 2010 edition of AeroSafety World there is an article by Michael W. Gillen on a study of the performance of 30 pilots flying for U.S. carriers on 5 standard manoeuvres (pp 30-33). The mean performance on all 5 manoeuvres was below that required for ATPL (4 on a scale of 5). Two of the manoeuvres had a mean which was below that required for "basic instrument flying" (below 3 on the scale of 5).
A CTL manoeuvre was not one of the five. If you are running an airline's safety program, however, it seems you would be unwise to assume that all your pilots perform at least to minimal ATPL levels on standard instrument manoeuvres. You would probably be unwise to completely trust your flight crew's unanimous assurances that they are all capable of performing and competent to perform CTL manoeuvres, given the conditions under which they are generally used. You might well be advised to test them specifically, to be sure.
But of course testing uses resources, which might be scarce (money for salaries during non-revenue activities such as training and assessment; use of scarce simulation facilities, or money to hire them; use of training captains and other personnel). You might well choose to recommend your airline prohibit CTL manoeuvres in general, and only allow them to be performed at airports at which you judge them to be both necessary and less risky (for example, with no high ground or obstacles in the vicinity) by pilots who have repeatedly and recently demonstrated their competence at the manoeuvre.
Should you decide differently? What would be the argument to do so?
PBL