I have not read the literature about this "Jellet constant" but here is a guess: it's constant if you can neglect the energy loss due to friction (but if there is slippage there has to be some energy loss).

I agree, spin decay could easily be the broken assumption here, whether due to tip friction or aerodynamic drag.

Now that you mention it, the 2 articles I have that mention the Jellet constant treat the torques about the CM due to tip friction but ignore the effect on spin rate, which was probably assumed constant. Nor was drag considered in any way.

In fact, I've

*never* seen a quantitative treatment of top behavior that made

*any* attempt to take dissipative spin decay into account -- and not for lack of searching on my part. Not even numerical treatments. Worse yet, those admitting that they're ignoring spin decay due to tip friction seldom even mention drag, which is probably the greater braking torque in many if not most real tops at speed.

For that reason alone, we have to be careful about applying formulas and calculations to real tops. The only formulas that don't suffer directly from ignoring spin decay are the formulas for the minimum speeds for stable sleeping and steady precession. Unfortunately, other practical issues complicate their application to real tops.