I wouldn't draw the conclusion that supply is not moving towards the enemy units at hex 56,53.
1. Bombing does not impede the auto overland movement of supplies.
2. Bombing base infrastructure per se does not lead to increased consumption of supplies as repair of the damaged infrastructure does not consume supplies.
3. Bombing of an airfield can result on a direct hit on the supply depot and that will destroy some of the stockpiled supply.
4. Bombing of a port can also result in a direct hit on the fuel storage tanks (in addition to the supply depot) and that will destroy some of the stockpiled fuel.
5. The combat report shows a negative modifier against "fatigue". By far the most common reason for having high fatigue is the lack of R&R (in a well stocked base) brought about by constant fighting/movement, particularly in jungle terrain.
The conclusion I would draw from that combat was that your opponent rushed into battle units which had not been allowed time to recover from earlier exertions.
Ah, if I were the Generalismo in charge of AE development, there would be several things I would do differently but you are correct that I am loathe to criticise the developers for ultimately having a different approach. For the simple fact is that AE is highly abstracted, despite the veneer it gives of being detailed, and accordingly it is always a subjective decision how to handle any issue.
Furthermore everything is closely entwined, which most people who regularly demand a "fix" for a perceived problem simply do not comprehend. In his post #10 above LoBaron makes this interconnectedness well.
When you start thinking about the appropriateness of consuming supplies to effect infrastructure repairs, you quickly come across the huge limitations imposed by the abstracted concept of "supplies". In a recent thread discussing festung Palembang, Bullwinkle made the correct point that the real cost in repairing infrastructure IRL is not the food consumption but concrete, steel girders etc, all of which are represented in this game (and yes it is a game and not a simulation notwithstanding what others say exactly because of these abstractions) as part of the output of Heavy Industry. So if you start to think about costing infrastructure repairs you would need to not only consider expending supplies to allow the engineers to feed themselves as they work but also the expenditure of Heavy Industry points to pay for concrete and steel.
Once one starts to go down this path of trying to "realistically" cost this activity, one also has to revisit the existing expense incurred when expanding Japanese industry. It would be a perverse outcome to find that filling in craters in already built concrete runways ended up costing more in supplies and HI points than expanding an aircraft factory.
Here is another point. If we are talking about trying to make it more realistic, why should the consumption of HI points come from a global pool, requiring no explicit transportation to the location (and therefore prone to enemy interdiction) whereas consumption of supply points would come only from the location itself.
Of course the developers sidestepped all the above issues by their abstracted solution. In many respects their subjective solution is much more elegant than trying to "realistically" deal with the above issues. Not to mention the reduction in micromanagement inherent in their solution. More micromanagement does not per se make a game better, or more appealing to potential customers in general. Even if they "merely" limited themselves to going back to the classical WITP solution of consuming supplies to effect repairs, those who criticised that earlier approach would reappear on the scene now.
To get back to your specific question. Yes, just left to me, I would be in favour of imposing a cost on repairs. But, as always it is a much more complex issue and would necessitate the allocation of developer resources which simply are not available. I understand the reasoning why it operates as it does now and it has the huge benefit that it has a logic which is generally consistent with the other game elements. I'm not convinced that my solution would be any more elegant or more efficacious and in the absence of it being demonstrably so, we are left only with subjective criteria. Which brings me back to my fundamental point that when it comes to subjectivity, one should always allow the developers of a game the freedom to exercise the whole of their vision and not compromise it with inconsistent add ons.
As to Monte Casino, yes the rubble was a great benefit to the defenders precisely because there was no "fortification" present before the rubble appeared courtesy of the monastery bombing. Without the rubble Monte Casino already represented a difficult objective due to its fine location, the quality of the defenders, the lack of quality of the American unit and the overall lack of Allied infantry in theatre. Once quality infantry under superior generalship, combined with indirect avenues of attack rather than frontal attack was introduced, the battle ended. Rubble is good but no one will ever settle for it in preference to having prebuilt pillboxes with enfilade fire supported by preset killing ranges over open approaches for artillery fire.