Theres a side of the argument that no one has mentioned yet.
When diving tables, diver are cautioned not to dive a "square profile". I learned to dive with US navy tables and as part of the training it was understood that when used on a square profile (remaining at the max depth for the max NDL at that depth for the whole dive) the tables had a hit rather of about 1% and this rose to 4% for repetitive dives. As a result we added in fudge facotors - next deeper depth or next longer time on hard working or cold dives and next deeper depth and time on hard working and cold dives, etc. We also normally took care to avoid staying at the max depth for the whole dive and rounded up - a 61 ft dive always became a 70 ft dive.
When you look at multilevel tables (PADI Wheel, etc) or at a computer one of the first things you should notice is that the the NDL's are much shorter, even when compared to newer tables based on doppler ultra-sound, Bulhman tables, etc. The reason for this is that the computer calculates depth and compartment loading every few seconds, so every dive is in essence a "square profile" and the only way to add in a safety factor is to reduce the NDL's.
But what that means is that the responsibility for adding in a safety factor shifts from the diver to the computer - unless the diver also knows enough deco theory to understand the limitations and the factors a computer may not consider, such as how cold you are, how hard you are working, how much you drank last night, your hydration state, how many dives in a row you have been diving, etc.
In my opinion, that knowledge is vital to using a computer safely.