Semantics, but per NFPA 79, that's not true (see below). Emergency Stops are discussed in a section there called "Control Circuits and Control Functions".
+1 for alien.
I know what P-J's getting at with his E-Stop design, but I agree with alien - GFCI devices have a certain lifetime as well, and I'd wager that it's at least a couple orders of magnitude less than a contactor. I did a little Google research, and I've found nothing to say that it's prohibited under the NEC to deliberately introduce a ground fault, it's generally considered to be a Bad Idea.
We had a whole thread on this a few months ago - unfortunately, it looks like you'd have to buy a copy of UL Standard 489, but it appears that UL certification for a GFCI breaker only requires a rated life of >50 fault trips:
https://www.homebrewtalk.com/f170/gfi-breaker-pricing-cant-right-295194/
A little other research turned up NFPA code 79, which is specific to industrial electrical design. You don't have to follow Code 79 in your home for sure, but I'm pretty sure that a commercial operation would/should be following this as part of their design.
The one relevant quote I found in there in talking about equipment grounding was:
(14.1.1 basically says that connections should be secure.) According to that reference, grounding conductors are only allowed to be connected for the purpose of grounding. Since the resistor/fuse trip method introduces a deliberate current to ground from a hot wire, that point where the hot wire is connected via fuse and resistors to the ground conductor is the specific violation.
Reading more of NFPA 79 suggests that the resistor/fuse method further violates their definition of an emergency stop, specifically requirement (3) of section 9.2.5.4.1.1 which states "Reset of an emergency stop circuit shall not initiate a restart". Most versions of the resistor/fuse design I've seen do not satisfy that requirement - you reset the E-Stop and the GFCI, and the power is automatically restored to your PIDs (and therefore your SSRs) and your pumps unless you switched them all off.
A modified version of Kal's safe-start addition to his brewery would satisfy the requirement, as the main power contactor relay can't be re-energized if any of the switches for PID power or pumps are in the on position - the only position that allows the contactor coil to be energized is with all of the pumps/PIDs off and the key switch on. In other words, it requires definite action on the part of the operator to re-energize outputs after an E-Stop. Another HBTer provided a simple working schematic for a NFPA79 compliant E-Stop here:
https://www.circuitlab.com/circuit/5zd3fg/holding-contact/ - combining that simple setup with the ice-cube relay and NC contacts on all of the switches from Kal's design would meet the requirements. Once you reset the E-stop, the ON button doesn't function until all of the other switches are off.
Another important point of distinction is that
an E-Stop is not supposed to prevent you from getting electrocuted inside a control panel. Lockout/Tagout (LOTO) is a separate issue. I work in an industrial food manufacturing setting, and the rule is that if you're opening up something electrical to do work, you isolate the power upstream and lock it out (on our systems, by putting a lock on the circuit breaker, local disconnect, etc). If you just hit an E-stop for a system, that'll prevent you from turning any outputs on. To open an energized panel after hitting an E-Stop, you either have to throw the main disconnect switch (which kills everything in the panel back to the point where the wires enter the enclosure, or if authorized to work in an electrical enclosure, use the bypass switch to open the door without turning off the disconnect (usually a screw that has to be turned while pulling the door open).