It is not a flaw as much as a cost benefit decision on Tarm's part. They used that detail up into newer models and as far as I know this is not a common failure. In my experience this is no stress riser if properly welded. When a weld is laid down it creates a molten puddle that would convert that sharp corner into a corner with a radius. The problem is that that only works if the person welding up the assembly had gotten a full depth weld so the puddle makes it down to the backside. I do not know what welding process they used at the factory but in general a gas shield wire fed MIG is lot easier and healthier for the employees. As mentioned, the trade off with MIG (or TIG) is less penetration than from a conventional welding rod based weld. It looks to me that they used some sort of punch press to punch out the slots. Thats quick and repeatable in a factory. The problem is the holes produced are going to be straight with no taper. The normal approach to get around poor penetration with MIG is to taper the hole out at 45 degrees down to only thin section that is straight. That would require a different process, in low production, an air grinder and burr. This would add a lot of time to the assembly, so the temptation is to just try to depend on the welders skill to get as much penetration as possible. Alternatively they do the extra weld prep and raise the price to the point where no one buys them. Tarm has a 20 year guarantee and they could make reasonable assumptions about how many times the boiler is cycled from hot to cold and make an economic decision that not prepping the joint is a worthwhile risk. As it is, Tarm's seem to last a long time past the guarantee so the trade off was worth it at the time. I have heard of earlier coal/oil Tarms having a large gasket leak on flanged connection. They tended to start with minor leak that was barely noticeable to most homeowners and by the time it got bad enough then mating surfaces were rusted and pitted and could not be sealed without building it up with weld and grinding it down to establish a smooth surface. That as the end of many. In most cases these were old units well past the guarantee. Ultimately any piece of equipment is compromise between commercial interests and design and ultimately if no one can afford it, it does not get built. NASA, in theory goes way to far on the design end while Space X goes with iterative "good enough". That works and cost's less until someone is strapped on top of one of those rockets.
Standard arc welding generally has far deeper penetration, the trade off is it requires more skill to keep from burning through the metal. Unlike MIG, anyone with a stinger can run a bead outdoors while a shield gas application requires a stinger and gas supply and if there is any wind or air flow past the weld the shield gas can get blown out of the arc. Since I do not currently own an arc welder, I use my MIG and when I need to weld big thick stuff for structural purposes, I do a lot of grinding to taper back the joints. The first weld can be hairy to keep the metal from burning back but after that I just do multipasses to build the weld back up to the original thickness. If everything is clean, I get very little slag or inclusions.
BTW I had to keep an eye on a project once to install a higher alloy tee into a main boiler steam line running at 850PSI at 875F (superheated steam), in that case the joints were preground and then they welded TIGed the root pass then spent the next 20 hours building it up with stick. There was a post weld X ray required and when they did it they found an inclusion half way down in the weld, the welders then needed to grind down the weld to the inclusion which took another 12 hours and then had to weld it all back up and Xray it again. Odds are if that inclusion was missed it would have been years before that inclusion propagated through the weld and caused a failure but when it did if anyone was nearby they would be parboiled in an instant if they were not hit by shrapnel. You will hear quite often about large construction projects that have issues with quality control, the temptation is to skip or fake Xrays as fixing one bad weld can ruin a schedule and cost tens of thousands of dollars. Odds are the people involved will be long gone before its an issue. In my case that expensive T was never used and that piping got torn out and scrapped less than 5 years later.
I guess it comes down to if you really want to go overkill on the repair, feel free to do so but IMHO, clean it up as well as you can and then get a good welder with fresh out of the package rod to run a bead over the cracks. Anymore than that is overkill. Note a lot of folks are cheap and keep rods around or take rods left over from jobs home. The problem is the coatings on the rod are somewhat hydroscopic and absorb moisture, if used in that condition, the moisture in the coating can lead to premature failure due to hydrogen embrittlement (crappy welds) that also can be prone to leaking. Ideally welding rods need to be stored dry in special ovens or hermetically sealed in packages.
