Sorry, but I do not follow you on this. What does it mean "better"? Less conservative is better? Closer to the US Navy tables? Or do you mean those advanced algorithms available on some AI computers which take into account the "effort", evaluating it based on air consumption? In this latter case, I can see the point of preferring an AI-equipped computer over a traditional SPG...No, you should follow what that computer tells you. It is good enough for those purposes. He is saying that in terms of the algorithm being used by a computer, you can do better.
Something, indeed, that I always did take into account also using the tables. Whenever air consumption had been larger than my usual SAC rate of 13 liters / minute, I always employed the table of the same maximum depth, but of a longer time, corresponding to the same air consumption I had in a normal-effort dive of this longer duration.
Example: my classical dive of 30 minutes at 30 meters, requiring a short deco stop of 3 minutes at 3 meters. The pressure at bottom is 4 ATA, hence my SAC rate is 13x4= 52 liters/minute. In 30 minutes, I typically use 30x52 = 1500 liters, which is roughly half of my 15-liters tank at 200 bar. Hence typically I still have 100 bars when I reach 9 meters and launch the SMB. Suppose that I see that the SPG indicates 67 bars instead of 100, it means that I had an anomalous air consumption, and instead of using 1500 liters (half of my tank) I had used 2000 liters (2/3 of my tank). I consider the extra air consumption equivalent to extra bottom time, and I follow the table for 30 meters, 40 minutes.
Of course my super-cheap Leonardo has no knowledge of the air in my tank, so he cannot make this adjustment to my deco stop. Does an AI-equipped computer makes the correction as I always did with the tables, or the pressure transmitter is simply used for sending an acoustic warning when the pressure falls below a threshold?
If this is the case, I really do not see the point of an AI computer.
If instead the knowledge of the pressure in the tank allows the computer to factor this in the deco algorithm, then probably the AI feature starts to become quite useful and important... And I understand why you say that a better computer can be safer than my Leonardo.