No. The plot is for the defined supply voltage. There is a less heat if you have 50W power with 30V power supply than with 48V. Considerably less.
So you're saying a lower voltage -- which would incur more current draw for any given power output, right? -- would result in a lower case temperature than a higher voltage, which would draw less current at the same power output? That doesn't compute here -- but the computer between my ears isn't programmed for electrical engineering, so you could well be correct. At any rate, running my two A07s at 48VDC has resulted in zero problems here -- their cases are barely even warm at idle and still hand-friendly (and much cooler than the power supply vents of my vintage Yamaha DSP-1 processor, which is rated at only 30 watts) at near-deafening volumes. Since one of my SMPSs has variable voltage output, I can informally test your contention by turning it down to 30VDC and seeing if the A07 it powers runs any cooler.
An hour+ at the same medium-high volume later: no discernible difference in the temperature of the top of the A07's case between 30VDC and 48VDC -- and IMO no real world reason to deny myself all that extra headroom.