I'd say look at roller test results for tires and adjust by a factor of 1.8 for poor quality chipseal, 1.5 for good asphalt, cement 1.4, unpolished wood velodrome 1.2 and 1.1 for polished wooden velodrome.

Keep in mind crr will not be constant over 40k as temperature and surface quality will change.

---

What's your CdA?

The best resource I can provide you with is Silca's work on the subject:
https://blog.silca.cc/...stance-and-impedance


In short, the quality of the roads you ride on can have a notable impact on crr and therefore speed at a given wattage.

0.004 is pretty good on a regular road. It's assuming very good tires, latex tubes, and a decent but not 100% perfect road surface, and a warm day. (CRR goes up in cold weather.) I've seen low 0.003s on a really smooth road with the best tires/latex tubes on a warm day, but that's unusual. As an example, a few days ago I measured the CRR for a decent but not stellar road racing tire (vittoria rubino pro) with butyl tubes at 95F degrees on a fairly smooth road and got 0.0055.

So it wouldn't be unreasonable to estimate 0.0045 - 0.005 on GP4000IIs (19mm rim/25mm tyre/85psi) with latex tubes on a very average road?

That sounds about right. It depends quite a bit on the actual road surface - smooth asphalt vs slurry seal vs chip seal etc, but yeah.