A typical lamp. Say 36W linear fluorescent. Or a 23W CFL. The same lamp from Sylvania, Philips or Opple. The same lamp in economy line and pro line.
Are there some differences in filament thickness, shape or material? (Wikipedia says they're "typically made of coiled tungsten" so it admits other possibilities...?)

Subquestions:

• Do cathode shields prolong lamp live or do they just prevent end blackening and lamps reach their EOL within the same time as without them?
• Are halophosphate lamps different that their switching cycle is considerably lower than that of triphosphors? I can understand shorter lifetime hours due to a different chemistry but why switchings?

I'm asking because some lamps seem to withstand real abuse such as many thousands of switchings while not preheated and other don't.

The effect of cathode shields depend on what is exactly their design. They may serve just as the shields and so just prevent the end tube blackening. Or they may serve as an electrode to take over some of the cathode (mainly the high speed ion bombardment at cold startup) or anode load.

Same purpose may serve different lead wire "extensions" (the lead in wires are longer and bed over to form a kind of rail in front of the filament, so take over the current when the cathode does not have sufficient emission).

Then the filament shape plays very significant role in where the fast ions end up: It is known, than the beehive like shape tends to redirect the ions to collide to each other and so dissipate the energy that way instead of "sanding off" the emission layer. This trick is common with LPS, but there is nothing, what would prevent the use of the same trick on fluorescents, well except such filament forming has some manufacturing cost.

The phosphor coat by itself has no means to influence the lamp life, except the gradual lumen depression of halophosphates. But lamp models with different phosphor mixes some usually from different model families, so differ in many other aspects of their design and itr is those other aspects, what may differentiate the lamp life. Different makers or even the same maker with different model lines may choose different marketing strategies with the introduction of the new lamps:
The "rare earth" is considered as "a higher quality" lamp with higher selling price, so may contain more features protecting the cathodes.
Or the rare earth lamp, being newer development the better knowledge allows better design of the rest, so improves the life without rising the cost so much.
Another strategy the higher cost of the rare earth "eats up" extra part of the expense budget, so the rest has to be more aggresively "optimized" to stay within the same selling price of the lamp.
Or with the rare earth introduction, the halophosphate were shifted to the "poor man't" category, so further cutting cost allows to keep the margin together with way lower selling prices.
Or the new lamp is designed mainly as "greener", so the use of more efficient rare earth based phosphor mix to boost the efficacy may come together with lower mercury dose, so earlier faults coming from "that direction".

So indeed, there are differences in lamp designs even when all are of the same designation (e.g. the F36T8).
To the previous, motivators for the changes may be the expected use of simple electronic ballasts, which heat up the electrodes only after the ignition, so needing more cold cathode robustness. Other makers may offer only the programmed start ballast, so optimize their lamps for these ballasts, so those lamps would then perform way worse on the cold start or preheat ballasts and so on.

Thanks for the exhaustive answer! It makes sense.

Same purpose may serve different lead wire "extensions" (the lead in wires are longer and bed over to form a kind of rail in front of the filament, so take over the current when the cathode does not have sufficient emission).

I suppose you mean a "filament support" with the filament attached somewhere halfway in direction from the cap to the tube, shaped in a special way.
But I'm not sure what moment (or what current) you mean by "take over the current when the cathode does not have sufficient emission"...?

The part of the current formed by the positive ions, when there is high voltage across the cathode fall.
When the cathode is cold, the closest conductive part with the same potential takes over the ions, so their beating. If that is some piece of wire or ring, does not matter that much, it is just the fact it is the closest thing in the discharge path, what makes it a (temprary) cathode. Then, after the filament with the emission coat heats up and starts to generate the electron cloud around it, this cloud then redirect the ions to the main cathode - although further away, the thermionic emission makes it the preferred "target", so then the lamp runs normally. That eans, when the cathode is still cold and there is high cathode fall, the emission coat is protected from the fast ion bombardment during the lamp startup.
This trick is usable only, when the filament heating up at start does not have to rely solely on the ion bombardment as the heat source.
So works only with ballasts heating the filaments electrically. But except the US "single wire per lamp end" style IS ballasts, practically all are heating the filaments electrically.

Whats the point in using cathode shield in only one end of the lamp as Ive heard some manufacturers and lamp models have done?

Whats the point in using cathode shield in only one end of the lamp as Ive heard some manufacturers and lamp models have done?

It could be not the real cathode shield, but a kind of "substrate" for the amalgam mercury pressure regulator. It is placed around one of the cathodes to get heated by that cathode, so the mercury pressure control become less dependent on the ambient temperature (a feature you will need for either wide range or high temperature or low volatile mercury lamps)...