KitchenBench - this type of approach preceded the VAR's. You went from beacon to beacon and followed a beam track with a variety of procedural turns. The so called beams had a morse code A and N and single tone when you were on the beam. Off to one side you got morse A off to the other you got morse N. Passage to different points was the various AM marker beacons you can see. There was no cockpit indicator like the VAR that came later.
This is the SFO approach that resulted in the BCPA DC-6 accident there long time ago. The Radio Beam navigation system was very common in the lat 1930;s and 1940's there were several transmitters on the main Brisbane to Sydney and Sydney to Melbourne Routes. The VAR system added the blue/yellow segmented display ala the old ILS gauges. From memory Tamworth still had the last VAR in the world up until the 1970s.
Checkboard - point noted however whilst I generalised the notion of any circling approach mostly being used at the completion of the final approach segment because of the result of winds (or cloud or vis) the point is this that you are at MDA and within the circling area and able to manouevre the aircraft to the base or downwind leg keeping the runway visually in sight and hence effect a normal landing, if I am only making a heading change and can pick up the final approach path within the arc of the runway ends then you are not conducting a circling approach. The fact that the MDA may be normal circuit height is not the issue it is the obstacles around the place and your ability to keep clear of them and the circling area is the only place that gives you a guaranteed 300 ft obstacle clearance. That is why your circling - you cannot land from straight in!
Anyway bugalugs with his Cat3 arrival should sit down and see how he'd go doing it a 747 - not! AS anybody who has done a few would know there are heaps of places where one side or the other is off limits for any circling at all, so it s a real squeeze before the weather and darkness makes it an even bigger squeeze.