We need to distinguish between shock hazards, which are proportional to voltage level in terms of the proximity required to cause a shock, and arc flash hazards

We need to distinguish between shock hazards, which are proportional to voltage level in terms of the proximity required to cause a shock, and arc flash hazards which are proportional to the available current and the speed of the nearest upstream protection device. Low voltage systems are often much more dangerous for arc flash than high voltage ones: the current is often higher (thousands of amps instead of tens of amps) and the protective devices (circuit breakers or even fuses) are slower. That means the amount of hazardous energy discharged into a worker before the circuit is automatically isolated can easily be enough to maim or kill even at relatively low voltages such as 120V. There's a 208V transfer switch in one substation I work at that has potential arc flash incident energy of 38 cal/cm^2. That's more than enough to kill even from 2 metres away. At 480 or 690V it is sufficient to simply not touch energized equipment to avoid being shocked, and PPE is available that allows hand-contact (class 0 gloves). However, available arc-flash energy in the event of a fault while the energized equipment is exposed can easily be above the level that any PPE can protect from (usually considered to be 40 cal/cm^2). If the arc flash hazard is at a level that cannot be mitigated with PPE, it will have to be reduced in some fashion either by re-configuring the circuit or de-energizing it before work can be done. Employers are responsible for performing engineering studies to determine the arc flash hazard level at every relevant location in their facility, labeling it appropriately, training employees on these hazards and providing adequate PPE. Unlike voltage, you can't just look at something and tell how much arc flash potential it has which is why engineering studies need to be done in advance. We have widely-accepted standards such as NFPA 70E and IEEE 1584 that codify the appropriate methods for engineers to use in estimating these hazards, and there's no excuse for failing to do that.