[...]
A Mercedes car goes faster in a straight line than a McLaren car, you can still conclude nothing about chassis-generated drag from that statement.
Perhaps the Mercedes overheats and blisters it's tyres in race conditions and must be configured to minimize cornering loads and compensates on the straights. Perhaps a McLaren works it's tyres more effectively and is configured for more downforce, gaining much more in the corners and over race stints with less tyre degradation or with improved tyre temperature in the damp, than it loses down any straight.
... So, in one simple jump, we moved from talking about speed traps to inspecting suspension geometry
The design efficiency and performance of competing wings will not necessarily be equal nor the performance plots be linear.
Perhaps a skinny McLaren rear-wing generates equal drag as an equivalent low drag Mercedes wing but generates less downforce for that drag and is therefore ultimately slower.
Perhaps a high downforce McLaren wing generates much, much more downforce than an equivalent high downforce Mercedes wing. In that situation Mercedes optimum solution would be to run low drag and be fast down the straight, McLaren would run high drag and be faster round the lap.
Again from this conjecture we have learned absolutely nothing about the relative drag generated by the respective chassis shapes.
Standing on the outside, without access to the race-engineer setup sheets or factory data, just looking at a facile speed trap number and then trying to argue one way or the other about radiator drag co-efficient is, and will always be, an utterly futile exercise.
And definitely not an argument worth getting worked up about.