Only that ... this types of performance trends suffers when the peak is not high enough... am not trying to diminish its presence on the track tho,...
I am really looking forward to see what 2020 would look a like for McLaren,...
How competitive McLaren will be next year will be interesting for sure. Going by Seidl's comments, we could roughly say that they hope to make as big a net gain next year as this did this year. That would be impressive to me. It would also suggest that there was a lot more "low hanging fruit" on the development tree than I had expected at the beginning of the year. They have made a few comments recently about how the MCL34 was a bit rushed - they could only start the development once they fully understood the problems with the MCL33, which happened very late going by the normal development timelines.
Talking of normal development timelines, from the few hints have been dropped, I get the feeling that it's roughly something like this:
12-18 months prior to launch: a small team looking at a variety of very theoretical models / car concepts
9-12 months prior to launch: based on winter testing and early season results of the current car, the various possible concepts are refined down to a single solution and the core properties of the car are finalised (things like suspension geometry, chassis dimensions, cooling layout)
6-9 months prior to launch: development starts to shift to the new car, starting with the bigger (longer R&D duration) components. They might run some early prototypes in Friday testing based on this.
3-6 months prior to launch: all R&D focused on new car and major components (eg chassis) would be in production
0-3 months prior to launch: initial development completed and only production and testing remains. R&D shifts to early season upgrades
So for the MCL34, they couldn't really do this due to the problems with the MCL33 - until they properly understood the problem it wasn't practical to commit to the design of the MCL34 because they had to be sure that they weren't repeating the problem. In short, the MCL34 is something of a "rush job" by normal development standards.