But, think about it, Tx is present for how many years? about 4? Since then we had the 2017 and 2019 iMacs with no Tx chip inside, only the iMac Pro got it which had a new design internally because of the components and the cooling system used.
Fusion drives may be a reason, but they could also just leave an entry model with it, just like now.
They delayed 4 years to fix the MBP keyboards, they forgot the Mac Pro for 6 years, the mac mini for many years too...
They didn't delay the fix to the MBP keyboards for 4 years. Apple was dogmatically stuck on making the butterfly to work for 1.5-2.5 years. It is only after they switched. When Apple makes a hard call on where the future is going , that's where they are going. Thinner notebooks are the likely the future Apple is moving to. They had to give a bit for now on the butterfly keyboard but that will probably result in them stripping out height some other way in the future . ( dumping x86 solutions will be part of that.)
SATA drives are the "past". The only reason the Mac Pro 2019 got some SATA ability is because they already had to buy the Intel I/O chipset (PCH) that had it anyway. So since paid for and there was "extra" internal space they added the sockets. But if there was no space it would have been dropped.
Apple Silicon is not going to natively support SATA. iOS and iPad devices don't need it so it isn't coming.
APFS file geared toward HDDs at all. It probably begrudgingly got some HDD and Fusion support but the basic operations of the file system are not geared to HDDs at all. APFS on HDDs is at minimum awkward. Selling entry level Macs with APFS layered on HDDs is selling pain points.
Apple has been on the "Flash drive is the future" track for more than several years now. If the Mini is dumping SATA then the iMac was inevitable.
When Apple Silicon Arrives internal SATA drives were going to disappear. The Mac Pro may be a very narrow corner case with an discrete SATA/USB controller added to he logicboard.
So, if the iMac is a dying product, because of the AS coming, why bother after all these years? Why offer a matt display option?
the iMac isn't going away. There were iMacs with PPC CPUs in them before. When Apple switched the x86 there was still iMacs. There really isn't a good upside in Apple tossing the "iMac" brand away in the trash can. Yeah they have mumbled something about no new "i" prefixed products, but the iMac as a brand name is not new. They could try to use simple "Mac" as a product but the major problem there is that "Mac" is pragmatically the name of the product line ( at least the short version).
Something like "asMas" would be silly. They have a good name for the midrange desktop Mac product. There are exceedingly few reasons to throw it away just because the internals are changing.
In general they paid too much attention than usual for this one, it is a bit strange...
It is not strange at all if they aren't going to have a 27' class iMac Apple Silicon solution for another 1-1.5 years. If the laptop range is changing more rapidly they are going to need to have something substantive to "hold" the desktop market in the mean time.
So pour all the other desktop stuff in 1080 webcam . 10 cores from iMac Pro. 10Gb/e from Mini and Pro . etc. That will help keep the Osborne effect from casusing problems during the somewhat lengthy wait to get to the other.
Then Intel transition was a "Big Bang" one that happened quickly. That was enabled in large part by two things. First Intel has a full can complete line up of x86 CPU packages to move to . Even more than Apple was going to every buy. It was a relatively huge line of CPU options from which Apple coud select a specific subset.
Second, they could lean on Intel to do some of the work to get the new logic boards across. Intel hands out "reference design" boards for the CPU production. Can have a generic working system up fast so all the individual system vendors have to do is built a board that is a variation of what Intel provides. Some rumors that appears on this site even hinted that Intel did most of the Mac Pro 2016 work. That is one reason why the board had more mainstream features. SATA routing about the same. Pretty standard ATX form factor. etc. Even more so leaning on the XServe.
[ Intel didn't have a good 64-bit laptop processor to shirt to. But they had something halfway decent if limited to 32-bits. ]
Both of those allowed Apple to under promise and over deliver. ( give themselves two years. And finish in about 15 months for vast majority of the line up and 18 to finish it all; non-GUI, single-user XServe trailing on the tail end. )
Neither of those are present here at all. Apple has no full line up of Apple Silicon to move to at all. They hare very little history of doing any such that at all. The bulk of their iOS and iPadOS products all operation on "hand me down" processors. When the do a new SoC it goes into all the new products that year. While iPhone 11 ( and 2020 SE ) line up. One single SoC. Watch? One single SoC. .. same exact configuration. iPad Pro ... 12X -> 12Z . Exact same die with on GPU "core" turned on in over a 1 year deployment. All of that is
not a vendor that is providing timely, concurrent, broad ecosystem CPU packages for a highly diverse set of products.
So pretty likely that Apple will do a gradual roll out of Apple Silicon. They will do a really good laptop SoC. Pause and then do a really good mid range SoC. Pause and then do a really good high end SoC. Then repeat the cycle.
And even if there was a full line up there is no broad spectrum "dot the i's and cross the t's" service to provide reference designs and run down a wide sprectrum of compatibility problems. Every new Mac on these first transitions is going to be pretty close to a "greenfield" move. No reference designs to "crib notes" off of. Yes, Apples designers aren't inexperiences. But Apple also tends to take having their own Silicon as a new opportunity to drive the product into a new "painted into corner" that the hardware now has to bail them out of . The combination doesn't lend itself to be timely.
[ The deeply pipelined and concurrent iPhone development process heavily masks the issue that those designs take a long while to come out. They aren't worked on sequentially. ]
In contrast, the Mac line up seems to have a "walk and chew gum at the same time" issues. if the MBP is being iterated on then the MBA can't change much. new Mini , no iMac. new iMac Pro , no Mac Pro. New Mac Pro ... no new iMac Pro . There are lots of indications that Mac have a finite set of resources allotted to them. ( some fixed subset of the industrial design , some subset of the OS core team , etc. )
So this time when Apple says " about 2 years" there is pretty good chance that will slide a bit over 24 months just as much as a bit under 24 months.
That will also give the folks who "have to have " SATA , m.2 slots , Bootcamp (bare metal) Windows/Linux , etc. time to adjust to what is going to disappear.
It could also be a sign that the iMac Pro is dead for good.
That may be in part more Intel and AMD than Apple having nothing to "jump to". I wouldn't bet on the iMac Pro disappearing. If there Navi 2 shipping now they would be something to move the iMac Pro to. Likewise if Intel had delivered what might be coming early 2021 in the Xeon W series ( an Ice Lake variant of Xeon W ).
With Apple Silicon there probably is still room for a iMac Pro. ( higher capacity and ECC RAM. Better screen , higher core count. ) that is the AIO price range just below the still ~$6K Mac Pro.
High probability that Apple will use Apple Silicon to move to smaller , 'slimer' systems . The Mac Pro is relatively just pain big in the Mac line up. Apple will be looking for something smaller and better literal fit on top of the desktop.