I have a simple(minded) idea here. Connect a damn horizontal rod between the two pistons/moving-legs on each side
Are you a programmer? Do you like it when a non-programmer tells you "just do X", because it must obviously be so simple?
Why, then, are you so sure that you can come up with a better design than mechanical engineers without even studying the problem?
It just feels like these ergonomic goods producing companies are selling chairs/tables with super high markups ($900-$1500) for ergo chairs - really?). Seems like you have a better handle on the production costs of a hight-adjustable table. How much does it really cost?
Making shit is expensive unless you're doing it at scale, the majority of people don't care about adjustable desks so the marketshare is by nature a niche. You then have staffing costs, design costs, distribution costs.
In the latter case, you would usually synchronize their controller clocks, and specify the motors/cylinders/whatever to both have a very tight positioning tolerance. Then you mount the surface being moved on a system of bearings that have enough freedom s.t. it will self center automatically, absorbing the tiny bit of misalignment that can (and WILL) occur.
However, in my experience it's almost always better to use two passive, i.e. non-lifting, supports along with one actuator in the middle that is twice as powerful. In heavy lift applications it often comes down to getting a single cylinder that is powerful enough without blowing the cost through the roof. Nevertheless, if you look at marine architecture, airport gangways, earth moving machinery etc. you will often see just a single lifting point.
And in three paragraphs we've gone from office chair pistons to front-end loaders! Yay HN! As I said... for what it's worth ;-)