The thing is, in motor racing, you're always trying to push the boundaries. It's an arms race, but in terms of chassis and aero. If everyone thought the same way, we may as well go to a single spec chassis. Part of the problem, as I've said up-thread, is a lack of testing. Solutions that look good on paper (or a simulation) may not necessarily be so good in practice. I've experience there, looking at the correlation between simulations and and practical applications.
As far as the Merc floor is concerned, it's a good way of generating downforce. As the speed increases, so does the downforce, which makes the edges of the floor form a "seal" with the track surface. It works much like the old side-skirts used to do on the original ground effect cars. The issue there is that it seems to be *too* effective at times. Again, as I said up-thread, the relationship between velocity and downforce is squared, so a small change in velocity will lead to big changes in downforce.
Christian Horner is in danger of the old "what goes up, must come down" because Red Bull have their own technical problems as well. DRS failures and fuel feed issues. Max is already down an engine - something he's conveniently glossing over right now.
Compared to BOTH the Red Bull and the Ferrari, the Merc might have its pogo-ing problems, but, touch wood, the car's been pretty reliable so far. This could potentially be a massive advantage for them later in the year when other teams around them are going to have to start taking penalties for components. It's no point having a fast car if it keeps breaking down.