Many thanks to Les Allwood for crafting this historical introduction
As ATC developed over the years, there was an obvious need for Reporting Points, to define routes and to ensure procedural separation. Beacons of various types defined many of these. The beacons were named after their location, and a two or three letter abbreviation chosen for the Morse code designator. For example, the Morse identifier for Strumble is STU.
Where beacons did not exist, a suitable geographical name was used, and the two or three letter designator was taken from an obvious abbreviation (often the first three letters). Sometime in the early eighties, the proliferation of navigation databases made this impractical as, without international co-ordination, many countries could be using the same designator in fairly close proximity.
The potential for confusion was obvious, and ICAO decided that all non-beacon Reporting Points would be given a (pronounceable) five-letter designator, and that such a designator would occur only once in any ICAO region. Over a period of some months, the old UK Reporting Points were given their new five letter names, almost without exception a recognisable facsimile.
So far, so good. Then, some years later, the Paris office of ICAO, which was responsible for the allocation of designators for new Reporting Points, made a strange decision to actively allocate designators that had absolutely no connection with a geographical feature. This was the era of XIDIL, NUMPO, NIGIT, DIKAS, PEPIS, etc.
Happily this state of affairs did not last very long, and we are now back in the position that designators are, wherever possible, related to a geographical place name. Some of them are rather odd, but that is an inevitable consequence of the rapid proliferation of new routes and points right across the ICAO region.
Your assistance with documenting more information would be much appreciated, either using the feedback facility, or by e/mail.