To add some more detail - 

Buckingham certainly operated to a timetable and not "just" a sequence. A version of the WTT was published in the Appendix to the Peco Publications book. This was supported by the automaton that operated the storage/turntable yard - the "Automatic Crispin" (AC). 

The AC was in three parts - a timetable sequence punched into card - rather like a Jacquard loom - with a timing track as one of the tracks - a speeded up clock that advanced the sequence and would stop and "buzz" if the operator at Granborough Junction did not keep up - and finally a home-made disc-based block bell and instrument simulator. The clock was visible to all the operators for timing their own departures.

Also Peter explained in an RM article his version of wagon dispatching which used colored dots on the wagon soleplates which were associated with differing daily destinations. 

This was a truly pioneering system of operation - which has not been surpassed in sophistication to my knowledge even today - something like 50 years later. 

Now the issue of running a "realistic" WTT at exhibitions but not being boring seems pretty simple to resolve to me - in the 12 inches to the foot world there were exceptions to the published WTT to cope with unusual events on a daily basis.  So, exhibitors could add extra trains around a simpler WTT based on declaring special events that keep the systems busier and still be completely prototypical. My own system is set in North Wales coast, so for me, the prospect of an Eisteddfod or a choir competition, a market day, a heavier shoal of fish being landed, the racehorses being taken to and from race meetings along with travelling support, the annual works outing from Northern mill towns, or many others are all excuses for a busier-than-usual pattern. I'm sure football specials (FA Cup), fleet week, royal visits, etc would apply elsewhere. 

Adam

PS I am personally on a path to emulate Peter's approach at Buckingham - but using computers. I already have a basic WTT mapped out by grafting onto the real 1960s WTT for the North Wales system. I intend to teach JMRI about block signals and so achieve a "Virtual Crispin". An advantage would be that it should be a lot simpler to change the timetable.