All of what you say is correct, but if you want 100% assurance that the power has been disconnected from the motor, there is absolutely no question when you unplug it.
The next best would probably be a local disconnect with a viewing window in the disconnect so you can see the status of the knife blades. I have seen Canadian equipment with this type of disconnect. Not sure if their code requires the viewing window. Also probably not available for use in classified areas.
However local disconnects are not very common in the industrial plants in my area. Most equipment is fed from a remote MCC without a local disconnect per the exception to 430.102 in the NEC.
Also they do make connectors up to 600 amps, but they are not load break rated and they are very heavy.
Yes I agree unplugging provides the best assurance of isolation. However, in too many cases it is impractical and costly therefore it may be difficult to regard it as sensible solution let alone a universal solution.
Visible break isolators have been popular in recent times particularly in mining applications. Again they seem to offer a simple form of isolation as well as being local to the equipment. However, they have their shortcomings as well. So much so that we know of several sites that have deployed VBIs who are now looking for an alternative. Problems cited include:
- Difficult to determine whether the disconnect is open - yes that's right the very problem that they claim to address. It can be quite unclear even for electricians what to actually look at inside the disconnect. What chance does an operator have?
- Chemical / pollution causing glass opaqueness
- Hard to see at night
- Limited load break / fault make capabilities compared with conventional isolation switches
- Very high capital cost
- Two extra terminations per motor yielding high cost
The arrangement "equipment is fed from a remote MCC without a local disconnect" is very common in our neighborhood as well.
The more I see of the problem of isolation verification for mechanical works the more I question whether regulators have an understanding of the impracticality of existing approaches. Your Try-Lock-Try approach looks OK so long as the infrastructure is already in place. Yes it still is not the best for reasons discussed earlier meaning some people that have had to decide recently whether to implement the Try Start method and spend money installing Try Start stations everywhere have chosen alternatives.
I find that a lot of people tend to fall back on a Live-Dead-Live check even for mechanical works. I guess they regard it as a high integrity result. After a while they realise that the productivity of their electricians is very poor. That was the motivation for the development of a device that allows a non-electrician to do a Live-Dead-Live style check.