Thanks for your reply.
What is your position in regard to training for events/performance where gross efficiency is the primary, or a primary, concern (e.g. RAAM, etc.)? Similar to the above, or would the establishment of baseline values have more (if any) relevance?
Gross efficiency is no more and no less important in something like RAAM than it is in a match sprint. In all events, the goal is to convert the energy stored in ATP into external power, and hence forward motion.
In any case, while cycling efficiency appears to be mutable, unlike with running it is hard to point to any specific way of training that might be effective. Thus, once again it becomes train for performance, and let your physiology sort itself out.
Why is that incorrect?
On my phone, so not sure of your reference, but on the issue of replacement, I meant during exercise (as shown). Obviously if you want to maintain your body mass, you need to consume as much energy as you expend over the long term.
Thanks for the input and discussion guys! I’m sure putting an accurate number there has many variables, I was more aiming for a rough-ish estimate to get a feeling for the throughput, much like xert and that connectIQ App do. It’s probably not compelling enough for TR to look into - or is it? 
Cheers!