Run a couple of alligator clips to a lamp socket on a coin door. Takes 2 seconds.

I do it all the time. It's not hard to avoid shorting the thing. This method is also a nearly irreplaceable quick way of testing to see if a dead lamp is a problem with the matrix, the bulb, or a crappy socket. Anyone who has ever had to swap out 75 sockets on a Bally does this.

When I do it I clip a new socket into the circuit and swap bulbs in and out of that.

If the difference between flipping your breaker and not flipping your breaker is some little bulbs then you really need to reassess your circuit load. 80% of total capacity is the guideline. Breakers are not foolproof and can be physically restricted by stuff like corrosion and rodents.

Pop Bumpers are a different issue. They aren't burning out. They are getting knocked out. LEDs will solve that but for a different reason.

In science we don't make assumptions such as an LED bulb lasting 5 years when they haven't even existed in the current form for 5 years. Yeah, an LED is not going to fail, but how about the resistor in the base? How about the contacts on the base? How about the connections inside the base? None of those things exist in an incandescent bulb.

LEDs themselves are proven to last longer than incandescents. The problem with all of the assumptions I keep seeing is that we aren't installing bare LEDs. We're installing a small device with multiple new failure points. Why do you think so many of the new LED bulbs are dead right out of the box? It's almost never the actual LED that is the problem. It's the manufacturing of the overall device.

The future problems are not the same once you have changed the operating model. Those earlier problems happened because the machine was on up to 100 hours/week. In the home the game might be on 10 hours a week. Once the damage from past problems is repaired there is no part of the home use model that will lead to those problems reoccurring.