Well, first of all using steel equivalency is a rather bad idea. The interaction between each armor type and penetrator type are unique, trying to shoehorn the provided protection into one single value vs all different types of APFSDS ammo doesn't work. There are research papers that have shown the flaws of RHA equvaliencies, the company Rheinmetall has stated - publicly and also in documents not available to the public - that RHA values are useless to represent modern armors, APFSDS ammunition isn't even tested against RHA anymore unless a customer demands it.
The values from Paul Lakowski are very questionable, I wouldn't rely on them. The assumed armor layouts, weight values and armor thickness values are completely off, most tanks are better/worse protected than others, simply because he assumed that the armor itself will provide more protection per weight.
That said a lot of older documents from the Cold War were completely focused on RHAe values. One document from the CIA/US Army, which has been declassified via the Freedom of Information Act, mentions that "one version of the M1 turret is rated as 750mm RHA against chemical munitions and 400mm RHA against kinetic energy munitions." Based on the age of the document (M1A1 had not entered service yet), this means that the data should be based on the basic M1 Abrams. Rheinmetall included a graph with armor and penetration estimations in one of its presentation, which put the M1 Abrams' armor at 350 mm vs KE (not mentioning if this is the turret or hull; also not mentioning the angle of impact). S. Zaloga has used the same value in one of his older books, also stating a protection level of 700 mm RHA vs CE.
During the late 1970s the British Army was developing a tank known as the MBT80, which was canceled in favour of adopting the Challenger 1 (that was originally meant to be an export tank designated FV4030/3 "Shir Iran" for Persia). There are documents from a computer simulation comparing the armor protection of the MBT80 and the XM1 Abrams. Based on the photos of the 3D models used in this simulation and the age of this simulation, it appears that the XM1 Abrams was the late FSED prototype from Chrysler, which was identical with the first production model of the M1 Abrams.
The simulation estimated that the probability to penetrate an XM1 with a KE round is 73%, with an ATGM is 50% and with an RPG is 25%. Hull down this drops to 53% (KE), 23% (ATGM) and 17% (RPG). The MBT80 was better protected with a simulated probability of only 65% penetrating the fully exposed tank with a KE round, but 58% and 30% with ATGM and RPG respectively. The hull of the vehicle was apparently not well armored against shaped charge weapons, because in hull down the probability of penetrating the MBT80 was only 28% with an APFSDS round, 21% with an ATGM and 11% with an RPG!
How much this says about the Challenger 1 is hard to say, as it wasn't a related development.
The Leopard 2's frontal armor has an areal density of about 3.5 metric tons per square metre according to German sources, which is equal to about 450 mm thick steel. It is not specified wether this is the average weight (suggesting the hull armor might have slightly less areal densit, while turret armor has slightly more), maximum armor weight or whatever else. The Leopard 2's turret armor package weighs 8 metric tons. Given that composite armor is designed to provide more protection per weight (the Leopard 2's armor has been claimed to include high-hardness steel, ceramics and/or tungsten by different sources), one should assume something in the area of 500-550 mm RHA vs KE.
This figure also is supported by other sources. In a book from 1986 written by Paul-Werner Krapke, originally leitender Baudirektor (~ project manager) of the Leopard 2 development, it is claimed that the Leopard 2 frontal armor resists a 125 mm APFSDS from 1,500 metres distance (for comparison, the original M1 Abrams was designed to resist 115 mm APFSDS from 800 m); this would again point to something in the area of ~550 mm RHA vs KE based on the penetration data of existing Soviet APFSDS ammo, and German APFSDS rounds (120 mm DM13, 120 mm DM23) that could have been used to simulate 125 mm APFSDS rounds.
The Leopard 2AV was designed to resist a 105 mm APFSDS round from 800 m (believed to be a round fired by the smoothbore gun of the Leopard 2K prototypes with 38 mm core diameter) and the MILAN ATGM with 600-650 mm penetration. The series Leopard 2 had better armor (at least in terms of coverage), which was kept until the Leopard 2A4.
On Soviet protection levels, there are multiple estimates available from declassified CIA documents, aswell as old Soviet/Russian documents. Harkonnen has published nice articles on his websites in the past, you need Google translator or other software to them however. In general the protection level of T-64A/B, T-72A and T-80B is believed to be in the area of 450-550 mm RHA vs KE and 500-600 mm RHA vs CE.