There is a straightforward explanation for the champagne stopper phenomena noted in your (article/item). In brief, they are due to the decline of resilience brought about by prolonged strain, and to differences in the degree of stress imposed on the corks. They are probably not caused by variations in the physical properties of the corks. Agglomerate cork stoppers are a manufactured product and are fairly uniform in their physical properties. There may be differences between manufacturers, and also differences of 1-2 mm in the diameter and length of the stoppers selected by winemakers.

My guess is that the observed shortening of the stopper can be attributed to a bore that is narrowed or constricted by fractions of a millimetre. This resulted in a lesser depth of insertion, a greater degree of longitudinal compression of the upper part of the cork, and a greater degree of radial compression of the stem of the cork. The greater compression of both the head (axial) and the stem (radial) of the cork increased the rate of decline in resilience and in the long term, caused a reduction in both the final overall height and the sealing pressure. (Further explanation below)

Inherent variability of the stoppers is not a factor and the major variables are as follows.

Controlled variables:   Stopper diameter and length, ‘leg length’ of muselet, nominal insertion depth, nominal gas content of the wine and nominal bore and finish dimensions for different bottle styles.

Uncontrolled variables:  Actual dimensions and profile of bottle bore, actual insertion depth, actual gas content of wine and temperatures during storage and distribution

Because a cork consists of some 80% air (by volume), it provides a pneumatic seal. After compression in the neck of a bottle, the elevated pressure of the air in the cork cells prevents the ingress of atmospheric oxygen, and at the same time, this pressure causes the air to permeate very slowly out of the cork to the atmosphere; the greater the degree of compression of the cork, the faster the rate of permeation and the decline in resilience. Thus the rate of decline in resilience of sparkling wine corks is faster than that of corks for still wines.

A ‘champagne’ cork, (75% air), is compressed in two directions. The business end is compressed radially from 30 or 31 mm to a nominal 17.5 mm, and its volume is reduced by at least two-thirds. The pressure of the air in the cork cells is then of the order of 900 kPa. The upper section of the cork is compressed longitudinally between the muselet and the rim of the bottle by the brute force application of the muselet. The extent of this compression depends on the depth of insertion, the effective ‘leg length’ of the muselet and the external dimensions of the ‘crown/cork’ or ‘cork’ bottle finish. Minor variations in the profile and dimensions of the bore can affect the depth of insertion of the stopper, and its sealing pressure.

Another variable is the amount of gas in the wine. In my indirect experience with ‘méthode champenoise’, bottle pressures before disgorging were about 800 kPa, and about 600 kPa after disgorging and liqueuring, (other companies may have different specifications). There were no further formal checks on bottle pressures, but after ten or more years, gas contents were noticeably lower. The need for retention of adequate carbon dioxide in the wine for a decade or more must be a consideration in decisions about initial bottle pressures and stopper dimensions.

Like corks for still wines, champagne corks vent when the internal pressure exceeds the sealing pressure of the cork, and the seal is restored after the drop in internal pressure. On several occasions, I have seen ‘complaint’ bottles of sparkling wine that had leaked twenty or thirty mL of wine but were still retaining 400-500 kPa of gas without any leakage. The most plausible explanation being that they had been stored at >40°C, the cork vents liquid, (one had leaked/vented on the back seat of a parked car), the pressure is reduced, and after returning to normal temperatures, there appeared to be an inexplicable loss of liquid from a sealed bottle of unspoiled sparkling wine.

The resilience of the wad or gasket in metal caps also declines over time, but there is not much quantitative information about their performance; long shelf-life is not a feature of other products. The nominal shelf-life of bottled beer is 9-12 months, but this is largely due to oxidative decline. My own impression is that very old bottles and cans of beer tend to be a bit flat.

Stelvin wads from old bottles are heavily indented, and the rate of loss of SO₂ in wine under Stelvin for several decades is about the same as that for cork. The performance of both closures depends on the nature and dimensions of the original materials and the quality of application.

I’m inclined to agree with Ed Carr’s comment about the amelioration of the flavour that occurs with some wines under cork. It also seems possible that the current vogue for ageing finished sparkling wines could be due to the lack of exposure to cork before disgorging; that is, because of the use of crown seals for tirage wines over the last few decades or so. The widespread belief that ‘ageing’ of wines under cork is due to the ingress of small amounts of oxygen is incorrect and not supported by any evidence. The circumstantial evidence suggests that it is caused by cork’s affinity for volatile, hydrophilic compounds in the wine. “Breathing corks” was consigned to the scientific rubbish heap about seventy years ago but strangely, it still has a strong hold on the hearts and minds of otherwise sensible people in the wine industry.

Despite being permeable to gases, corks provide an impenetrable barrier to atmospheric oxygen because the concentration/pressure of oxygen in the compressed cells is much greater than that in the atmosphere. However, carbon dioxide is water soluble and, theoretically, it can permeate through a compressed cork. Although there is significant loss of carbon dioxide from sparkling wines, there is no post-bottling oxidation, except for wines sealed with plastic stoppers.

Differences between natural corks for still wine are nullified when they are compressed to 50-60% of their volume, (picture yourself squashed into a 40 litre drum!). Their sealing pressure depends mainly on their diameter and that of the bottle bore. The difference in appearance and performance of natural corks extracted from bottles of wine is not due to inherent differences between the corks, but to whether or not they have been subjected to high hydraulic pressure after insertion. Winemakers would like you to believe that any problems with ‘oxidation’ or so-called “leakage” are caused by inferior corks, but this is just a ‘cop-out’. The fault lies with the generation of headspace pressure during or after corking; this results in a softening of the cork cells, and also provides the motive force for the expulsion of wine past the cork. It’s a bit like carpenters complaining about bad nails that bend, go in crookedly or take more than the usual number of blows to drive home.

The erratic behaviour of your ‘leaking’ bottle is difficult to explain. The sealing pressure of the cork and the gas content of the wine both decrease over the years. During this time, there is a dynamic, quasi-equilibrium between the gas pressure in the cork, the sealing pressure, the gas content of the wine and the gas pressure in the bottle. When the bottle was removed from storage, it is possible that the movement disturbed the equilibrium, and the internal pressure exceeded the sealing pressure shortly before the wine was cooled. I was told by a hands-on ‘champagne’ maker that the way to test old sparkling wines is to shake the bottles close to your ear. Bottles with little or no pressure make a ’gurgling’ noise. If the gurgling noise persists, the bottle is flat, but if the gurgling stops, there is still adequate gas in the wine. I have never had the need to apply or test this.

I quite like old white sparkling wines with low gas content, but the sweetness of sparkling reds can be a little cloying when there is very little gas left in the wine.