The phrase DR tank is frankly ludicrously amusing concept.
Yeah doing more math on it, it really is kinda a hoot. The more I look into it the more bad of an idea it is to do anything but the bare minimum (I can't stress enough I do not endorse the 'swimsuit runner' plan, armored jackets do make a meaningful difference!) is a bad idea. Just get as good a initiative upgrade you can and cheese out full defense, which as a bonus works surprisingly well even on extremely high PR goons, while DR maximization basically stops working at all really early. It is comical how bad an investment DR is for not taking damage compared to just boosting athletics or giving yourself extra minor actions so you can always full defense.
DR has Nothing in Common with AC. Zero. Zilch. Nada. DR will never prevent any damage, ever. In the extremely unlikely even that your GM is kind enough to let you gain a point of edge and then not cancel your spending of it that edge might, possibly, prevent damage, but it still won't be the DR doing it. If DR did prevent damage then it would actually be useful and there would be no AR/DR problem.
That is a similarity to AC, in that it helps reduce incoming hit rate.
It just is AC is a meaningful 5% reduction in hit rate, and thus incoming damage on you, for each point of it, while AR and DR takes the non-intuitive sliding scale aspect of opposed pool rolls (Which have many advantages, but a super clear exact understandings of your probabilities for each point of investment without knowing the context of the roll isn't one of them) but turns it up to 11 by making it not even resemble a steady drop in expected gains but instead fly about all over the place in value.
AC I know is a 5% reduction in hit rate. Period. Boom. Done. 5% reduced expected average incoming damage, we are all square h- well actually the fact that 5% of damage is going to be crits changes that and variable crit thresholds and damage multipliers changes that but its easily 'good enough.'
I know a defense dice is anywhere from like 2% to 12% lower a hit rate, with the number trending lower the more soak dice I have. I know that (At least in 5e) that due to how the skill dice scale, I can expect the average professional fighter to roll around 8 dice with more elite enemies using more and less elite enemies using less, and with enemies getting more unpredictable in behavior and capabilities as they go up. That is still a lot of uncertainty, but it is 'predictable' uncertainty. I don't know how I stack up to a given random prime runner but, knowing about what prime runners I might expect to face lets me have a good aproximation, and even if I am 'wrong' my dice investments still have value.
AR and DR is the worst kind of unpredictability because its maximum uncertainty: You don't have real relationships between AR and DR and pools on NPCs, AR and damage, a concrete way to evaluate DR or AR on NPCs based on grunt rating (it does generally go up but it gets really unpredictable REALLY fast), and the outcome of investments requires you to make the 'perfect' chocie to matter, so guessing wrong either means ALL your investments were worthless, or whatever 'extra' investments were worthless.
The payoff for 'correctly' managing defense and soak is you don't take damage, which is critical in SR as fights tend to be about objectives rather than survival or grinder combat in 5e, so taking a chunk of damage significantly hurts your odds (As it has serious consequences and tends to either floor you or put you close due to how soak worked). With edge, you can guarentee 0 DV attacks which is critical when getting proned by a big hit (Even if you aren't KO'd) means you can't do something critical like stop a runner or prevent someone from hitting the shutter close button on the building your in or whatever.
In SR6, correctly managing AR and DR gives you or denies the enemy one edge, which is significantly weaker than in SR5. Your really pushing the results by about 2 dice, which can matter but it doesn't matter a lot, it will almost never make or break a fight when SR6 fights are also longer which makes even slight combat advantages more valuable and you don't need to push your advantage as high as you do in SR5, or worse, SR4, where fights are over almost instantly.
So the system is less understandable and coherent, says less about the world because it doesn't track to anything, and is lower stakes in terms of the payoff but higher stakes in the sense that even slight mistakes significantly reduce the meager reward of it at LEAST. I know you can't linearlly compare systems like this directly but I think its fair to say that AR/DR is 'worse' than soak in almost every way save for exactly one thing: Attacks that hit almost always deal damage but don't one hit down, allowing SR to emulate a dungeon crawl where you grind resources down which... I don't think is what SR should be trying to be, its a heist game at its heart, but eh, different strokes, I wouldn't kick a spinoff edition focused more on capturing the crowd brought in by SRR looking for more grid based cyberpunk D&D if it came up with cool systems and encouraged awesome fight designs like SRR (sometimes) did and took it as an excuse to re-balance all the roles along a combat axis. Call it Shadowrun: Tacitcal Ops or something. Heck, I basically tried to do a mega-paired down more grid focused SR system for a 'Urban Brawl' pseudo-boardgame hack. SR in part biases towards heisting because it works so well as theater of the mind, so a more fighty resource management mercenary focused variant wouldn't be unwelcome.
The issue is that plenty of games arrive at this equilibrium without almost completely gutting the combat mechanics of the system and looping almost everything into either being a AR/DR modification, or a new way to spend edge that isn't worth it because now you actually DO need to race to do damage and don't have time for interesting thematic combat choices.
