I am familiar with the work of Kargoz, having watched a number of his vanilla class guides.
And yes, you're right that the damage per mana (DPM), on the immediate untalented face of things, LOOKS almost impossibly bleak for Arcane ... to the point where you'd have to wonder "Whiskey Tango Foxtrot would ANYONE EVER consider using any arcane spell for damage ever??" ... and truth be told, most people looked at the raw spell DPM numbers like you've referenced and simply never looked back.
But doing that leaves out things like my discovery of "shaping" WHEN you spend mana on your spells (so as to recover mana from Spirit while continuously casting), which then acts as a huge NET mana cost reduction over time, while at the same time keeping damage production "up" over the baseline when doing a cast/channel/cast/channel/repeat rotation like I'm advocating for in this build. This pattern also helps you with your endurance for long duration combat rather markedly, to the point where it starts feeling a bit like you've got a bottomless mana pool, especially if you've been stacking on as much Spirit as you can on your gear. So it's a very different philosophy for how to reach for and achieve success.
But it's when you start mixing schools of spellcasting like I'm advocating for here that you start finding some unexpected inter(re)actions between talents and spell schools that start making stuff a bit more complicated than simply looking up numbers on a chart and plugging them into a spreadsheet. Needless to say, I'm still in the process of testing out all the theorycrafting on what I'm doing with this build, but so far it's all panning out really nicely, to the point of being better than I'd been expecting (honestly).