The statement regarding the hull armor package can mean anything and is hardly relevant. The armor package of the Abrams has been upgraded several times - like the adoption of first, second and third generation DU armor in the turret - but this was always limited in scope due the limits set on weight and budget. This is why I mentioned that the M1A1HA (from 1988) had the same turret bustle armor array as the M1 Abrams from 1980 - maybe the materials were improved, but any gain in protection purely by material changes is likely very limited when looking at NERA's performance vs KE) - and even modern Abrams' tanks have the same side hull armor as the original production model (at least in terms of thickness and layering, again small adjustments could have been made to the materials). Likewise "new hull armor" might be identical to that of the previous model with slight changes in material composition; it could be different, but weight and size impose limits.
These factors also make it rather unlikely, if not flat-out impossible for the Abrams to resist 125 mm APFSDS rounds at the hull (specifically when looking at the frontal arc rather than just the frontal surface; 65 mm skirt armor + 60 mm hull armor is not going to stop much).
The design of ammunition can be altered at any time of a product life; before, during and after initial introduction. Svinets-1, Svinets-2 and Relikt might be very different in terms of internal construction or material composition since then.
You keep relying on baseless speculations. First of all it is extremely unlikely that any modern APFSDS performs worse against sloped targets than against flat armor; sloped armor is easier to penetrate and perforate - by as much as 20% in case of 60° sloped armor according to the work of W. Lanz and W. Odermatt (depending on penetrator geometry).
The performance of APFSDS ammo against steel armor is also completely irrelevant to their abilities to penetrate composite armor; research has shown that the ability to penetrate special armor is highly dependent on the exact interactions between armor and projecitle; two rounds with the same penetration against steel armor can have very different performance against the same type of special armor, which was demonstrated in the German tests during the LKE program (two rounds with the same penetration against RHA had a difference of 110 mm in penetration power after defeating a special armor array). This also means that all figures in regards to armor protection (and penetration) have to be taken in the context of the munitions or simulation used to come up with the values: an armor providing 500 mm protection against a BM-22 APFSDS won't provide 500 mm protectioon against something like the M829A3 APFSDS round.
You also cannot simply say "this armor is newer, so it has to provide significantly more protection". You want the armor of the later models of the M1A2 Abrams to provide 50% more protection, while staying in the same physical volume, retaining the same basic steel shell (as old tanks are rebuilt and the interior and exterior steel plates are apparently not changed unless necessary), roughly the same weight limit (may or may not be different for the M1A2C), being affordable and providing decent multi-hit capability. Specifically given that supposedly (assuming the front and back plates of the turret have the same thickness as on hull) more than 150 mm of the "600 mm" protection that the M1A2 (HAP/EAP) provided against older types of APFSDS ammo are the result of the steel structure, you are asking for an improbable, if not impossible, improvement in performance.
Armor doesn't grow on trees, neither does it come from the clouds of imagination. MBTs achieve higher levels of protection by adding weight and/or volume to the armor. Improvements from superior materials are often negligible and come with another issue (large increases in costs!). The M1A1 HA's much higher protection than the M1A1's (using inaccurate RHA values apparently an increase from 380-400 mm to 600 mm) came in combination with an increase in armor weight by roughly 3 tons. The Leopard 2's massive gains in protection came with an increase in armor thickness by up to ~80% and a weight increase of 5 to 7 tons depending on model. The M1A2 SEP series doesn't seem to offer similar increases, although there are still some open questions regarding the M1A2C (How much of the increased weight is related to armor? Does the listed figure include TUSK or Trophy?).
The Russians claim that Relikt provides a reduction of APFSDS penetration by 50% even against APFSDS rounds designed to defeat Kontakt-5. You should note that your quote doesn't mention Relikt or the ability to defeat a T-90M/T-90MS/T-90SM anyhwere, which is relevant. The ability to defeat ERA is not a binary metric; improving the performance of an APFSDS round agains a certain type of ERA can mean anything from "We managed to reduce the performance penality caused by ERA by 5%" to "The penetration performance of the APFSDS is essentially unaffected by the ERA".
The text speaks of third-generation explosive reactive armor, which is a very broad term and may have absolutely nothing to do with the exact layout of Relikt. How would the US Army know about the exact working mechanism, layout and performance of Relikt? Super-spies like an American version of James Bond? Did the US Army simply call Putin and asked them about these things? Or did they go to Amazon.ru and order a bunch of Relikt ERA tiles?
Most likely the US Army uses a self-developed ERA system as representation of future/current ERA arrays; Germany did something like this during the LKE II program. The USA might have by pure coincidence developed a perfect clone of Relikt, which could perfectly replicate the behaviour of Relikt - but that is extremely unlikely. The US' third-generation ERA might look and work completely different from Relikt, for example it could be based on Nozh (a third-generation ERA to which the US military actually has access, because it was installed on a number of T-80UD/T-84 tanks purchased by the US Army). Nozh and Relikt use completely different working mechanisms, so a performance gain against one type of ERA doesn't automatically render the other type obsolete.
Threat vehicles is likewise a very broad term. It could mean that the US expects the M829A4 to defeat a T-90M, but it also could mean a dozen other things. Being able to overcome third-generation ERA potentially without a major reduction in penetration performance doesn't automatically mean that all tanks equipped with this ERA can be defeated; there still is some hefty amount of base armor, which supposedly has been improved on the T-90M/T-90MS/T-90SM, that needs to be penetrated aswell. Being able to defeat threat vehicles with third-generatrion ERA also could refer to upgraded last-generation tanks fitted with Relikt (or Nozh/Duplet), which might be immune to the current M829A3 (they should be immune, if the Russian performance claims were true). For example it might refer to the T-72BM or the initial model of the T-84, which should have inferior base armor compared to the T-90M.
We don't know if the M829A4 is capable to defeat the T-90M with Relikt ERA, we don't know if the current M1A2's armor is capable of resisting hits with Svinets-1/Svinets-2 - yet you keep making generalizing statements based on nothing but speculation. That is not good.
No, we don't know anything about the hull armor being improved. CBO reports are mostly based on unclassified data and use publicly available sources. Damian just likes to ignore any weak links in his sources as long as they fit his narrative. Just look what's under the table that Damian considers a confirmation of his theories:
Yes, another unclassified CBO report from 1993 and a privately-run website run by Gary W. Cooke...
Unless the United States decided to change the definition of heavy armor two times (before and after the AIM upgrade), the M1A1 AIM doesn't feature heavy armor in the hull. First M1A1 AIM tanks were made in 2000, but in 2006 only five prototypes of the M1 Abrams featured heavy armor in the hull.
As of 2014 General Dynamics was only granted the licence to install DU armor in the turret of the Abrams, but not the turret. The CBO report simply doesn't have the same degrees of quality and reliability than the documents from the NRC. Unless the CBO report is using the term heavy armor to refer to one of the five prototypes with DU armor in the hull or has changed the definition of heavy armor (which also seems unlikely given that the M1A1 AIM weighs 62 metric tons vs the M1A1 HA's 61.3 metric tons), it is simply incorrect. The only variant that theoretically could have heavy armor in the hull - and that depends on the weight distribution - seems to be the M1A2C Abrams.