I'm going to withhold judgement on this on until my spring cleaning. I don't run on auto, ever. I will shut the stove off during the day during shoulder seasons and fire up as necessary at night. Stove is in the basement living area, so it's not mandatory that the lower level is always heated. I'll burn just over 2 tons this winter (very mild) and I don't plan to clean my exhaust until spring. I only have 8 feet of pipe and it's 4 inch, so I shouldn't have an issue with lack of cleaning. In the spring I'll look at the buildup of ash and see if its acceptable. If so, I won't change to auto anytime soon.
To me, it just doesn't make sense to let the stove cool during the heating season, even if the weather is mild. Keeping the stove warm and dribbling a few BTU in the room while the stove is in maintenance burn doesn't seem like a waste. BTU is BTU. Only possible argument is the fact that the stove might burn less efficiently on low than medium or high. Maybe, maybe not.
Furnaces used to be on/off, now there's 2 stage and variable speed systems to keep trickling heat into the living area as it's lost. This prevents the system from cooling resulting in lost efficiency.
Based on my math, 4 bucks a bag for pellets, .75 lb feed rate on maintenance/idle would require 1333 hours of idle time to consume enough pellets to eat up a 100 buck igniter. That's 55 days of idling continuously. I think it would take at least 2 seasons or more to break even, not counting the 70 ish watts of additional energy consumed during each start up. So, 14 starts would consume 1 kW at your local rate.
How fun would it be to build a 500 square foot room inside a climate controlled cold storage facility. Then we could really test all of our different operational methods and hypotheses under strict process control. Hmm, Federal Grant???