The simple answer is you do get ettringite at the interface between render and plaster.
There are various reasons why this might not cause a failure. The first and most important is the level. The ettringite chrystalls are formed when moisture borne sulphate ions come into contact with calcium alluminates in the
cement. The level of formation will depend on the amount of sulfate ions present. Now if you think about a wall. Firstly it is generally drier and the general direction of moisture will be from the render into the brickwork. If moisture runs the other way ettringite will form but it will also be self limiting as the water pressure will be away from the render thus forcing sulphates way from the interface. That is untill there is saturation. Secondly the weight of plaster is generally light and so gravity is unlikely to "pull" the plaster away from the render unless there ismsignificant attack which you might see with a risng dampor penetrating damp problem.Thirdly you don't exert impact and dynamic loadings to a wall.
Now let's look to the floor. The direction of moisture movement will be from the gypsum based screed toward the
cement thus the concentration of sulphate ions is higher. You walk on the floor thus exerting dynamic and impact loads which will break the weakened bond at the interface. Other influences also come into play such as underfloor heating.
If you want to see an ettringite failure on a wal I shall take a pic of my kitchen wall and post it for you. The plaster is falling off the wall as I have a penetrating damp issue which I need to deal with.
The way to stop ettringite is to eliminate one of the three elements needed. The first being
cement, the second being the potential contact between the
cement and the third being moisture. These are dealt with by either removing the
cement and using non cementitious adhesives whichis the most robust method, removing the contact by priming ideally with an epoxy primer or thirdly remove the water. This last is the hardest to do as there will always be some moisture present.