What that plot is showing is that at one specific cell temperature/Battery voltage combination, PWM is as efficient as MPPT, at every other cell temperature/battery voltage, PWM is significantly less efficient than MPPT. So if your battery is nearing full (14V) and the cells happen to be at 70C, then PWM does as well as MPPT, however if your solar cell temperature is at 25 C or 85C or your battery is at 13V, PWM is 10 -20% worse than MPPT.
The efficiency of a decent MPPT controller is actually pretty high - 98% for the Victron units that are popular in these parts, and a buck converter is at peak efficiency when Vin is close to Vout so at low panel votlages the controller should be hitting this mark. I don't know what typical PWM efficiency is (ie the Rds(on) of the MOSFETS), but it is likely about the same. The power consumption for the controller itself (again based on Victron numbers) is 30mA at 12V, or about 0.36W.
I hate to call you out on this, but the 'reliability' point is also a red herring. If you are worried about this, there is absolutely no reason you can't wire your panels in the same fashion for an MPPT charge controller.
The efficiency of a decent MPPT controller is actually pretty high - 98% for the Victron units that are popular in these parts, and a buck converter is at peak efficiency when Vin is close to Vout so at low panel votlages the controller should be hitting this mark. I don't know what typical PWM efficiency is (ie the Rds(on) of the MOSFETS), but it is likely about the same. The power consumption for the controller itself (again based on Victron numbers) is 30mA at 12V, or about 0.36W.
I hate to call you out on this, but the 'reliability' point is also a red herring. If you are worried about this, there is absolutely no reason you can't wire your panels in the same fashion for an MPPT charge controller.