The original link contains 3 places on the tree for "WLP019/WLP051/WY1056". I do NOT take this to mean that these 3 yeasts are identical! Far from it! I think they just weren't sure which one went where.
Exactly - Suregork et al did a remarkable job of guessing most of the White Lab strains based on a handful of database hits and then just spotting patterns in the countries of origin - California/England/Germany/Ireland is a slightly weird combination of countries but it perfectly matches WLP001-4 for instance. But there were problems with a bunch of sequences described as from "California" or the US.
There were a couple of unidentified sequences from California and a couple of Californian WLPs that hadn't been placed, so it was just guessed that the two "gaps" corresponded somehow. Unfortunately White Labs didn't include them in the strains they unblinded in response to Suregork's work - the thought is that they may have actually sequenced variants that were banked by local Californian breweries and so were covered by confidentiality clauses. So we don't know for certain about those ones, but most of the other identities we can be fairly sure about.
But the picture is becoming clearer on the main US strains involved in the
Ballantine's ->Siebel BRY-96/97 -> Anchor/Sierra Nevada
chain. Bear in mind that :
Strains get mixed up, labels get switched, beer gets contaminated, stuff happens.
If Ballantine's was _anything_ like British breweries of the time, it would have been using a blend of at least two strains in its ale fermentations, but may well have been using just one for its lager.
Siebel would presumably have purified them down to single strains.
The West Coast breweries could easily end up with multistrains, either deliberately or by contamination.
Yeast evolve over time. Not hugely, but enough to be different at the DNA level and eventually at the beer level.
So what is the DNA telling us?
First - only WLP051 seems to be a lager strain, the rest seem to be ale strains. If WLP051 came from Anchor as is claimed, I'd guess it somehow came from Anchor Steam and not Liberty.
The latest sequencing puts Lallemand BRY-97 in the Mixed group - ie a distant cousin of Windsor/S-33 and nothing to do with the main group of brewery yeasts that includes the Chicos and Whitbread B. There needs to be more analysis of exactly how it fits in though. Since Lallemand own Siebel, presumably they got it direct from the banked version.
The main "Chico" group seems to have a single origin and is most closely related to WLP025 Southwold, presumably from Adnams. Adnams are interesting, they got their yeast from Morgans in Norwich in 1942 after an infection, in the 1970s it was 5 strains that they reduced down to two after stability problems. But the relationship isn't close - the split could have happened 300 years ago if you believe the Gallone timings (which I don't entirely).
WLP008 East Coast ("Sam Adams") is a distant cousin of the Chicos which could have split off in the mid 19th century.
The remaining Chicos are much more tightly related, they might have split from each other within the last 30-40 years, some more recently than that. In that group you have WLP001, US-05, WLP090 San Diego and Pacman for certain. I don't think we have something that definitively ties 1056 to that group but I would expect it to be in there - it's subtly different to WLP001 but not that different.
Although I don't have a definitive source, the Chicos are commonly claimed to be diploid (ie two sets of chromosomes like us) but
Pacman for one is definitely tetraploid (four sets). Doubling the chromosome count is something that's quite "easy" to do and seems to be quite common in domestication, particularly in yeast.
We don't have sequence data for 1272, all we have are some basic interdelta "fingerprints" from Richard Priess which suggest it's close to WLP002 and WLP007 in the Whitbread B family. It's worth emphasising that this can only be tentative even with perfect data and in this case the gels are a bit contradictory, but the family does include "clean" yeasts like WLP029 Kolsch and WLP030 Thames Valley so it's plausible. We can be reasonably confident that 1272 is an ale yeast, which means it's nothing to do with WLP051. But the identification in the Whitbread B family is not certain, there's maybe a 25% chance it's something else, which opens up the possibility that it's similar to BRY-97 but it's not probable.
An intriguing option would be to do some DNA work on the East Coast Yeast strains that supposedly come from Ballantine - ECY-10 Old Newark Ale and ECY-12 Old Newark Beer are meant to be the ale and beer strains respectively. Unfortunately they only seem to be produced sporadically, and ECY don't distribute to Europe AFAICT - I don't suppose anyone happens to have some in a fridge/freezer/beer? Given the huge differences between Chico, BRY-97 and WLP051 it should be possible to distinguish them using pretty basic DNA techniques.
It's worth emphasising that although this stuff is really interesting from geek and historical points of view, you shouldn't sweat it too much from a brewing point of view. As others have said, it's brewing performance that matters. But DNA can open fresh perspectives - like using WLP051 California V Ale to make a classic low-temperature pilsner, or adapting an APA that you normally make with 1272 American Ale II to use the better-floccing WLP030 Thames Valley instead. Neither of those would be obvious things to do if you just looked at the names, but make perfect sense from a DNA point of view.