I point it out because without knowing literally ALL details, modeling it, doing an FEA, or some flex test we're basically talking standard textbook engineering stuff.
If you take two parallel members, connect them with one crossmemeber welded at a 90 degree angle with no gussets, when you put a load on this (remember it has a big moment arm from the wheels and also the lower control arms are not between the mounting points on the vehicle) as shown below and it parallelagrams on a level that you probably can't see with your eye, and this also happens to a lesser extent to the entire chassis (which is why also triangulation as simple as the 73+ struts from the firewall to the fender supports helps enough that I could also feel it in the wheel on my torsion bar car). The tie bar you show will help somewhat IF and only if its not a rod end and it has no bushings. In a quick search, I couldn't find a good enough picture to determine which type it is. If it's anything but fully rigid it likely does a'most nothing. It's going to be minor against parallelagraming because again there's no triangulation.
View attachment 1716197453
In addition, looking from the front of the car, you have an induced torisional moment into the tube which is made worse by the lever arm below the k-frame. So only the wall thickness reduces it because you generally have one bolt and mounting point with this lever arm and it's not spread out across the entire support frame in the way it would ideally be. This is my point with the narrow pivots.
View attachment 1716197456
View attachment 1716197457
For an example, I own a 2004 Chevrolet Colorado ZQ8 truck (was replacing the wheel bearing here but this is the best shot) which is further lowered using adjustable coilovers, this has a front steer rack, SLA suspension, front mount sway bar and a coil over shock with a tower from the factory that does not turn, so generally the suspension design in the HDK / Alter-K / QA1 kits, but in a factory design setting. You can see the front and rear pivot points for the LCA which takes all the load are as far apart as possible. This spreads the load out, reduces dynamic flex, and provides better tracking overall. Each side is mounted with a double shear. The strut rod in the stock system is the second pivot point and it's controlling for-aft movement due to the triangulation as well as side to side and the moment. This design is basically if the metal is flexing under the moment it is what it is.
My own feeling having driven that vehicle over 100k miles vs my Duster with 1.06 T-bars, hotchkis non-adjustable shocks, etc is the on-center feel is better mostly due to the front steer rack and pinion, despite having more body roll than my Duster. I think the limits are lower, but it objectively feels slightly better at anything that isn't right at the limit. This truck also has the ball joint type sway bar end links which are more responsive feeling because there is zero flex.
Regardint the sway bar, there is never a perfect side load, it's always a tension load up/down on those points so it's unlikely it slides at all. But there is a load with an angle and that design will be under bending and torsion as a result of it being unsupported. It just seems like they would be flexing all over the place. The design in the Alter-K is different and IMO better in this area. So you end up with a torison moment around the cross car axis and bending with an angle relative to vertical with fairly minor support. Also in nearly equal and opposite directions when mirrored across the car centerline
View attachment 1716197462
Just for a comparison sake of what's out there, this is a Maximum Motorsports K-frame for a Fox Mustang. You'll notice it has one crossmember on the front, the cars are front steer rack, though they are strut, but it does have an A-arm. You can see they have triangulation tubes and gusseting in several directions and a tie bar that is solid that mounts between the rear body attachment points. It also has a much wider and double shear pattern for the lower control arm mounting and as far as I'm aware this has quite heavy wall thickness. The rack is mounted directly to the cross tube and the triangulation is nearly right at the mount. This is about as good as you can do given somewhat similar constraints.
View attachment 1716197463
I would say also if you want to talk about things being flimsy and not falling apart, the fox mustang, especially the convertibles is a fantastic example of something that has the chassis stiffness of wet spaghetti stock but doesn't fall apart. Not breaking is a minimum requirement. On a small enough scale, the entire car is flexing.
Also please don't take this as criticizing the quality of any fabrication regarding the HDK / Alter-k-tion / QA1 because I don't view it as dangerous, but the engineering side of it leaves quite a bit on the table for a handling application, and I'd say it's quite good on a drag car. I mean you could literally make some of these improvments after the fact if you like (gussets, support bars, different sway bar mounting tab design). Anything is possible.
The stock K-shape is just better in torsion at bending because the K is making its own triangulation in one direction and it's a large cross section box tube in the other direction.
