Tighter isn't always better. It really depends on the application.
I some cases tighter is better (within reason). The Scubapro Mk 10 for example is a great design, but it was designed for 3000 psi pressures. With a DIN conversion kit they *usually* work just fine at 3600-3800 psi pressures encountered in cave diving. However, that's not universally true. The piston stem o0ring seals the gap between the hole in the first stage body and the piston stem. If the tolerances between them are too large, the piston stem o-ring (90 duro hardness) will start to extrude into the space between the too and when that happens it can get pinched, which clips off a cresent shaped piece of it. Iy you remove one and it looks like it has feathers on it - that is what is happening. Leaks are then inevitable and you'll see bubbles coming out the ambient chamber holes. Over time with lots of used and wear, even a tightly tolerance Mk 10 can wear to the point that you start getting o-ring pinching even at 3000 psi. It takes a long time and a fair amount of neglect or abuse for that to happen, but when it happens, the first stage is history as Scubapro no longer sells new first stage bodies for the MK 10.
Tolerances can also change with age and/or use. For example, the spring on the Mk 10 tends to get stiffer with age (not strictly use) and after about 20 years in service they started encountering Mk 10s with springs that were so stiff that the IP could not be brought within range, even after removing any shims. The solution then was to replace there spring, but if the "new" spring in your parts bin (or in Scubapro's warehouse) had been made 20 years ago, that might not solve the problem. Scubapro instead designed a new seat for it, and then included 3 seats in the service kits that were in three different heights. Thus, if the spring is too stiff, you just install a higher seat, which reduces the spring compression and the IP.
Sometimes looser is better. I recall a conversation with Lamar Hires from Dive Rite regarding O2 solenoids on the Optima. Dive Rite was was sourcing them from a supplier who built them to their requested tolerances, which were on the snug side, on the premise that tighter tolerances would be a good idea in an O2 solenoid. However, after they started having problems with stuck solenoids, they figured out the tighter tolerances just increased the potential for a small particle of sand or silt to cause it to stick. They started specifying looser tolerances and the failure rate went way, way down.
In short, they wanted AK 47s rather than AR-15s when it came to that particular application.
---
When "tight" versus "loose" matters, you need consistency. You want the parts made to very close minimum and maximum dimensions so that you can control the relationships between the parts. That will ensure proper interchangeability of the parts without hand fitting, and it will ensure you don't run into problems when parts that interact with each other are made to opposite extremes of their tolerance ranges.
As indicated before, I've encountered some reg brands and models that have those issues. You can fix those issues - if you have access to enough parts to find parts that work well together.
