Part of the problem in discussing flicker is that people's ability to see flicker of various frequencies varies greatly. So a report of "bad flicker" or "no flicker" depends on that individual unless the measurement was via photodiode and an oscilliscope!
Edit 1/28: Since flicker sensitivity generally decreases with increasing frequency, it's not surprising that the faster RPM engines have less problem with flicker than the 650 rpm engines. The increased effectiveness of the flywheels at higher rpm has also got to help. So there's one point (maybe important if lighting is a big issue) for higher rpm.(end edit)
Most frequency meters are going to be looking at a time weighted average, either from the meter movement itself. So just because a meter says's the average frequency is within 0.1 Hz doesn't mean that there isn't variation of voltage/light output at 10.8 Hz. There most certainly is.
Even at just 0.5% modulation at 60 Hz,, there are folks who will be bothered. Often these are the people who subjectively "don't like computers", or "just don't like LCD displays". (LCD's are internally refreshed at 60 Hz regardless of the computer's video frequency and have 0.5-1% modulation.)
At lower frequencies like 10.8 Hz,, the sensitivity to light intensity fluctuation (modulation) is typically even greater, and I wouldn't be surprised to see some people notice even 0.1% modulation at that frequency. This level of modulation of peak voltage you would certainly see from acceleration of engine/generator at power stroke time.
As to the "beat" between the 10.8 Hz engine and the generator's 60Hz output: When the increasing velocity at the power stroke coincides with peaking output wave of the generator, you will get an abnormally high peak voltage. This beat will be a well below 10.8 Hz, and I don't know if it's an issue. Edit: Since this beat is sub hertz and the generator power pulse rate is so much higher, I think I was off base on this.