I was largely shielded from this reality at National Lab--we needed to have 2 papers a year, but that isn't too hard with appropriate levels of collaboration. Merit increases were definitely tied to having high impact publications, but since the timeframe was year to year, I just published when I was ready in the "best" journals that would publish the results.
I am finding this kind of depressing. I was always the type of researcher who mocked the "least publishable unit/slicing the salami" style of publication, but now I can really see the temptation. I can see the changes in my own work already--there is some data we have now that we are writing up as a communication. If I were still at National Lab, I would probably hold it back for some additional experiments, but I am too worried about other groups working in this area publishing first to let it go longer. It is too risky to me to hold on for more data, since we have a full story already. I really wish this weren't the case, but there it is.
I need to have a good publication year this year, and that is starting to trump other considerations. This is one of the realities of the TT that I knew was coming, but is still upsetting. I am still in a good place, after all, I have data that is good for publication in excellent speciality journals, but it is harder than I thought it would be to make the call.