Scubapro (before they were paired with Uwatec under the JWA umbrella) used to have computers that were notoriously conservative in the 80-100 ft range, especially on repetitive dives, and that did not make them popular with divers who dove with more liberal computers.
I never had issues with either Pelagic hocky pucks (Aqualung and Oceanic) or with a Wisdom as both tolerated deep stops very well and would usually clear a few minutes of deco with a slow ascent and deep stop long before you reached the 10' ceiling.
In my opinion, a computer that stacks on more deco time with a 2-4 minute long 1/2 max depth safety stop is not a good computer to have. Some computers are just not intended for intentional deco use and are not deco friendly at all. In their zeal to save you from yourself they create psychological pressure to quickly ascend to the ceiling depth and consequently subject you to what is literally a bend and treat deco model. Not a good thing at all.
Reading some of the odder fine print in the manual might help you determine what is going on. The Sherwood Wisdom manual for example indicated that it used what were essentially US Navy table limits once you got into deco rather than the algorithm it normally used and if you missed that line, you missed all of what it said about a different deco model. So in deco mode it was very liberal compared to the Palm VPM (VPMa) and DPlan tables I cut and used for deco. That was a good thing as the computer always cleared before I did and served more or less as a bailout option if I had to surface quicker.
I still use my Cochran EMC-20H the same way (as a backup to a computer generated profile) although I have the conservatism upped to 70% to keep it semi close to the DPlan profiles I normally use as the EMC-20H is I suspect the most liberal computer on the planet. In my opinion you'd have to be stupid to dive an EMC-20H with the conservatism set to less than 30%. It is a good example of why it is important to have a basic understanding of deco theory and various models before relying on a computer, as both overly liberal and overly conservative approaches to computer modeling have their drawbacks and pitfalls.
In general, it is nice to be able to adjust the conservatism of your computer to match a buddy's computer to ensure you both stay in the same ballpark, but it can take several dives to get them more or less consistent with each other.