Marcush, I don't think those Kiel probe sensor arrays are being used to optimize anything. They are crude and have a huge cross section, for sure they will affect the flow.
They seem to pop up about once or twice per team, and then be gone, and they pop up mostly behind the exhaust. I think they are being used as flo-viz in an area where there are no surfaces. Flo-viz is not used to optimize, it answers the question do we have separation here or not? They know the sensor arrays are changing the flow somewhat, but it is mostly about finding where the flow (exhaust flow, easy to detect by temperature) roughly goes (for example, how much does it spread vertically at 120Km/h and 15000RPM?). For that they yield a nice (rough) 4-D map (vertical, horizontal, speed, RPM).
They use it once, see if the models are at least in the right ballpark, then they can continue to develop the models. As you said, a rough measurement to make sure that there is no phenomenon not included in their models suddenly popping up. If your exhaust plume were to go straight to the outside tire when you are above 16000RPM and below 100Km/h, because the exhaust trips the downwash slot (just a random example), neither wind tunnel nor CFD nor pressure sensors would tell you, but you would still be frying that one tire for 5 seconds a lap in Loews.
Wind turbines are cool, elegant and magnificent. TANSTAAFL!