OK, I'll answer as an engineer - and this is simply what I do, NOT an instruction for anyone else. Everyone has to make their own choices.
Backing off 'a bit' from a full mechanical limit stop open position for all except quarter turn style valves has been a standard practice through my industrial career, and I also do this with my scuba tanks. I didn't get this practice in school, but from my father who was a shift working refinery operator from the time he returned from WWII to when he retired in 1981. If you've ever had to apply a pipe wrench of 24" or greater size, much less such a tool with a 'cheater bar' extension, to try to get a valve that normally runs open to even start to close, you begin to appreciate backing off just a tad from the hard mechanical stop at the end of valve travel. Then, think of doing this with the valve and its associated container filled at pressure with something other than air - say hydrogen, or natural gas, etc. Nice dry gases make very poor lubricants - in fact, they have a tendency to provide a 'stripping' action. But even compressed air is dangerous enough from a contained energy standpoint that I would rather not have to take tools to try to get a valve designed to operate by hand to operate.
The mechanical engineers call this 'breakaway torque' if I remember correctly (I'm not a mechanical engineer). It takes more force to start a valve to travel off its seat than to stroke the valve once off its seat. Just like it takes more force to start a car rolling, than to roll along - friction. If someone has well and truly 'jammed' a valve against its mechanical stop, or forced it so hard it started to damage the mechanical stop, the 'breakaway torque' could be significantly different, as expressed by several folks here.
Matt, looks like you & I follow similar practice.
A 'half-turn' rule sounds like overkill to me, especially if the valve is of the quarter-turn design (90 degrees of rotation = full closed to full open) . . . like ball valves as an example, or plug valves as another example. I just back off the slightest bit so the valve isn't forced 100% against the end of its mechanical travel, whether on scuba tanks or other applications - even the water faucet for the hose to water the grass at my house, for example.
I don't know the history of the evolution of the scuba tank valve, but I'd like the quarter-turn type feature myself - you can tell at a glance from the stem position whether the valve is open or shut, and one can easily employ purpose-built plastic or metal 'locks' to keep in one position (safety shower and eyewash station potable water supply lines are one example where this is handy).
Think of the "J" part of the "J" valve design, without the 'reserve' function - just a 90 degree rotation for open or closed, period. Why wouldn't a 90 degree rotation scuba tank valve be functional? Might be OK at differential pressure of 500 PSI or so, but maybe too much torque when working with 3000 PSI or so differential pressure without having an overly long valve handle to get leverage. Probably why we have the style valves in common use today on scuba tanks.
I have seen a few 'vintage' tank valves on Ebay that employ a lever that is pushed across the valve body, but those seem extremely non-ergonomic, especially when working against a 3000 PSI neighborhood of differential pressure, so the short life of that design is understandable.
On pressure gauge failures, I've never personally witnessed one on scuba gear, but have experience with the phenomenon in the industrial world. Pointing away when first pressurizing makes LOADS of sense to me. I don't know of a good way to predict when my SPG might fail - but I'm open to any tips anyone else might have.
I don't get overly upset if someone other than my dive buddy (wife - also engineer) 'touches' my tank valve, but I deliberately stop and put it where I had it set before proceeding further when this happens.
Just my choice.