The following answer is taken from Wardale's response to a letter from Angus Eickhoff, published in Steam Railway #276 (see 'Articles and Letters' page of this website):
"Compound expansion for the 5AT has been considered - and rejected. The arguments for and against compounding are too complex to air here, but I am firmly convinced that for a high-speed locomotive such as the 5AT, simple expansion - using all the cylinder refinements that are now possible, but which are not common knowledge - is the right choice. Mr. Eickhoff states that the compound has left the simple to catch up - well, I believe the 5AT will show that it has now fully caught up, and indeed probably surpassed the compound for power generation at high speed."
The following is taken from the 5AT technical specifications Item 16 (Par. 5.1.): "The proposed design will define 'state of the art' for 2-cylinder simple locomotives, and may serve as a reference level to which the performance of all other types of locomotive can be compared. Analysis has shown no net advantage for the 3-cylinder simple type compared to 2-cylinder simple, however the 3-cylinder option remains open if the dynamic augment of a 2-cylinder design at the proposed maximum speed is unacceptable to Railtrack. Whilst the 3-cylinder compound type, with one h.p. and two l.p. cylinders and with the l.p. cranks set at 90º, might offer somewhat better thermal performance, more especially at lower speed, an overall thermal improvement sufficient to justify the extra design complexity and higher manufacturing cost cannot be guaranteed. The limited l.p. cylinder volume possible within the British moving structure gauge, with a conventional layout of the cylinders, is an important limiting factor on compound design and performance."