I thought that was funny, too. My guess would be the tips they have would undergo deformation, particularly after the first yard or so, when the bullet has just cleared the muzzle blast sphere and is plowing into the air at its highest velocity, and that it would be much worse for the FTX soft ballistic tips for lever guns. But that's not what the radar is showing them. A melting phenomenon would have to be due to the longer exposure to heat from air friction. This would soften the plastic and make it easier for the air pressure on the nose to deform it. When it slowed below where enough heat is being made by air friction to cause that softening, the tip could tend to return to shape. A lot of plastics do when you heat them. I would expect the BC's exact value to be the least of concerns in that scenario, as it should be repeatable. Instead, the deformation being asymmetrical and introducing eccentric wobble of the bullet flight path would be.
The other funny thing is their talk about constant BC and appearing to show a G1 BC. They are not shooting bullets the same shape as the G1 projectile so it would be odd for the G1 BC to be constant unless the shape were not very aerodynamic. Something's just wrong there. I'm thinking they actually found the drag function wasn't constant, so a BC using the projectile's own drag function with a non-melting tip made of wood or other material was not constant, but that they didn't want to have to try to explain that.
For those not familiar with Doppler RADAR, it's the same technology used by traffic policing before laser speed meters came out. The military has used it for decades to follow bullets, because the microwaves it emits actually bounce off the bullet base and return to the receiving antenna continuously, producing a frequency shift that may be accurately measured, allowing instantaneous velocity to be calculated at any point in time, and letting them follow the projectile's velocity throughout its flight and doing so with higher precision than an optical chronograph provides. Properly done, this can resolve fractional feet per second accurately.
The limitations to Doppler RADAR are that the antenna being off to the side of the barrel causes a trigonometry error you have to compensate for, and that the FCC doesn't allow radar powerful enough to work at great range without special licensing. My guess is that Hornady licensed their unit for a particular location for which they were able to show no interference with aircraft radar or other microwave devices would occur. There is a home version Doppler RADAR called LabRadar, but I understand it can only reach out 50 to 80 yards because the company could not get FCC permission to make its transmitter powerful enough to go out further, even though their prototype went a couple hundred yards. Also, that unit only reads out velocities at particular ranges and does not appear, in its literature, to produce a continuous graph. The maker says specifically that it's not suitable for determining ballistic coefficients. That put an end to my interest in it.
What Bryan Litz did was get a pair of chronographs that agreed with each other (Oehler 35P and CED M2) well, and set them up at long separation distances apart and took the risk of being able to shoot through them accurately enough at long range that he wouldn't damage them. This gives him pairs of velocities at a measured separation, and BC's that are average over that separation are calculated from the velocities at the two points. The JBM site has
a calculator for doing this.
You can do something similar. You just have to keep the precision limitations in mind. Litz said he's tested a lot of chronographs and found 15 fps is about the limit of accuracy on them, so keep that in mind. They resolve single feet per second, or even fractions in the case of one German model, but the absolute accuracy is not to be counted on to be any better than 15 fps. He likes the large Magnetospeed V3 for rifle muzzle velocity determination.
