In the end we are back to the confrontation of the Maxim derivatives vs the Lewis, I will try to sum up the points
]1) Mechanical design:
Parabellum, improvisation borne out of necessity. Lewis, best of all of the era, all things considered, even if it had its faults
I have often read the claim that the Lewis was more reliable, without any proof. If anything, there's plenty of anecdotal evidence, regarding jams in overwing fixed Lewis, and the defects of British ammunition.
I have come to the following conclusions regarding reliability of mechanisms:
Vickers slightly more reliable than the parent design the Maxim, the Maxim sometimes suffered from broken firing pins. If ammunition quality is not an issue, then both guns are quite reliable, though the Vickers robustness and ability to fire during hours without breaking down is exagerated, if the gun were that good... it wouldn't have a manual with 15 pages devoted to clearing stoppages!
On the ground, both MGs should be more reliable than the Lewis, if for no other reason that the simpler the mechanism, fewer things can go wrong. I don't know if it's the case with the Lewis, but gas operated guns tend to foul up with gunpowder residue overtime. The issues with the feeding method have been exposed so I don't repeat them.
Put them onboard airplanes, and then the picture changes. All guns, regardless of type, suffer from extreme cold at high altitudes. All of them suffer from congealing of the oil, and it can also cause steel to become brittle wich causes breakage of parts. I conjecture that the stress put on the gun mechanism by subjecting cold metal to the sudden stress and temperatures of firing might explain the instances of a MG firing for the first time at high altitude resulting in a few rounds being fired and then a jam, due to breaking of a spring or firing pin. I am no gusmith, so I am basing this on anecdotal evidence of gun malfunction in WWII in the winters of the East Front, and basic physics.
3) Weight and ease of handling: Lewis wins hands down. Lighter, better balanced (no drum hanging from the side), slightly more controllable in recoil due to gas operation
The Madsen also had its own problems, including a very low rate of fire and a small magazine capacity. The Bren belonged to a later generation of LMGs, and was the most outstanding weapon in the world in that category. Everything else was inferior to it. It remained in frontline British service (converted to 7.62mm NATO) well into the 1980s, and I read only yesterday that some British soldiers were asking if there were any in store which they could be issued with in Afghanistan, because they wanted its long-range accuracy and hitting power.
Well, this is wandering off topic, but probably the main reason is because the L85 Rifle was a stinking pile of junk, and problably its light machinegun version as well.
The Lewis gun never had a Hazleton adapter - it was a muzzle booster which could only be utilised by recoil-operated guns like the Vickers. 700 rpm was the maximum the Lewis reached (quite respectable for WW1), the Vickers was accelerated to 850 rpm, but that was a bit academic in synchronised applications.
I beg to differ. In the Lewis gun datafile, Harry Woodman only mentions at the end of the book that
experiments were carried out to increase the rpm of the Lewis led to the guns being tortured to fire at rates of 850 plus, the guns almost ripped themselves apart.
The point of such experiments is lost on the author as the gun was still stuck with the 97 round magazine
Digressing again, the answer is that a high ROF increases the chances of a kill during the split seconds the target is in the sights,Hence the quest for even higher rates of fire including the Hazelton booster and the German Siemens "electrical guns". The problem with this experiments is that they only addressed one of the problems, the punch, remaining unsolved the other problem, actually hitting the target. So if you can't bring the guns to bear on the target, a higher ROF is just wasting ammo, and if the pilot was skilled enough to get close enough to don't miss a target (within 50 meters), a higher ROF is relatively unimportant since even a single MG will do. Higher ROF or heavier punch in the form of heavier caliber bullets or shells have to go hand in hand with improvements in sights to be fully effective.
Dave Watts, (FlyxWire) in similar threads found that Hazelton booster was indeed used for the Lewis, but I think it had another name. He quoted the following passage from another Woodman's book.
Harry Woodman, Aircraft Armament: The Aeroplane and the Gun up to 1918 page 45:
The Hazelton Adapter:
The rate of fire of the original Mk.1 ground gun was between 450 and 550 rounds a minute and adjustments could be made by tightening or slackening the fusee spring. The early synchronizing gears (in particular the Ross gear) tended to slow up the rate and it was not until the arrival of the CC (Constaninesco) gear that any major improvement could be made.
However, Lt. Cdr. George Hazelton RN invented two speeding-up devices for machine guns, one for the Lewis
and another for the Vickers. In the latter the device was accommodated within the muzzle casing and consisted of a special sleeve and shallow conical spring which increased the force of the gases which pushed back the muzzle cap. The gases were held in the muzzle cylindre and could not escape until the muzzle cap, moving back, uncovered the holes in the muzzle cylinder. One Vickers fitted with this attachment was made to fire at 1,000 rounds per minute but this was far too high for normal operational use: the wear and tear on the parts was considerable and they were also liable to fracture. The 'muzzle booster', as it came to be called, was thus adjusted and modified until the rate was brought down to 850-900, which remained the standard for this gun.
To sum up, the Lewis was the most highly regarded LMG of WW1, and was particularly favoured for flexible mounting in aircraft because of its light weight and compact ammunition magazine.
Highly regarded because there was no contender. Compact? Yes, but the spool on the Parabellum is compact enough also, your statement about the belt inconvenience surely referred to the empty belt flapping around in the slipstream.
You discount the jamming of the drum with G-forces simply repeating that it served for a long time.
I hadn't thought of that, but thinking that a loaded 97 round is close to 3 kilos in weight, wich can easily be trebled in violent combat maneuvers and that the magazine is meant to move around instead of being a fixed magazine, I find it plausible it happened sometimes with wing mounted Lewis, since the gun is fixed to the airplane frame, and the stress is borne by the attachment of the magazine to the gun. On the other hand a gun in a flexible mount would tend to swivel and pivot as a whole on its mounting, so it wouldn't be an issue for in that case.
And finally, in your book, page 15, you note that
ROF in belt fed guns is affected by the drag of the belt installation, (wich may reduce as ammunition is used up)
This stands to reason, I have always been skeptical of the theoretical ROF of machineguns in ground use seeing footage of how the belt is jerked as the machinegun eats it, and why there's always a gunner assistant holding the belt to make it flow easier. In airplane installation this problem is solved, but I have to ask you, aren't spring driven magazines slower in feeding? I think a belt feed is a smoother operation than a magazine, but I don't know.
Anyway, wouldn't the big drum of the Lewis also cause feeding drag?