Yes, the actual firepower column is not part of the table. Yes, an abstraction such as letters could be used in place. However it doesn't necessarily follow that the Defences column is the best that you could get. Shooting at those same defences at short range should be better. Of course, maybe there will always be some sort of loss in firepower no matter how easy the target is. This is debatable, but let's assume true for now. Well, as pointed out, the actual firepower column is merely an abstraction. It could be replaced with letters. Therefore, since WB firepower is an abstraction, there's nothing to suggest that firepower 12 (or firepower class L if using letters) couldn't cap out at 14, 15 or 16 dice. It is only if you assume that each point of firepower represents 1 potential die of damage that there comes a cap on the maximum dice that a given firepower can represent. Then you could argue that you'll never achieve 100% efficiency under any circumstances, but you couldn't argue that the firepower column is merely an abstraction. So it's one or the other. If it's just an abstraction then there's no upper limit and therefore 90% of no upper limit makes, well, no upper limit. If it's not an abstraction then you could argue for the inefficiency.
However that is surely debatable. It is not beyond reason to suggest that it is possible to get all shots on target. It's also not unreasonable to suggest some loss, regardless of how plum the target is. That is to say, neither scenario is particularly unbelievable. However, the boundaries of the current table are a little unreasonable. An abeam Eldar escort at long range, into the sun (BZ 1-4), with a dust cloud between should be practically invisible. Similarly, it should be easier to hit a defence at close range than normal range, particularly into the sun (BZ 1-4). So since there's reason for the extra shift left and it's not forbidden, logically speaking, to get 100% accuracy then it is a reasonable suggestion.
If you really really didn't think it possible to get 100% accuracy then consider this scenario: All WB strength on all ships is doubled. The values in the Firepower column in the gunnery table are also doubled. So they're 2-40 instead of 1-20. Now you could insert 2 more columns between the defences column and the firepower column, representing increased potential accuracy under certain circumstances while still giving less than "100%" accuracy. That's just an example of how both the "inaccuracy" as well as "better performance" could be operationalised, not a suggestion.
As an experimental rule I would suggest including the firepower column in the chart, as well as reducing incoming firepower by 2 or 3 for each right shift beyond the furthest right of the table. To give an extreme example, targeting an abeam Eldar escort at long range into the sun (BZ 1-4) through blast markers would lose you 8-12 firepower.