My info came from a trusted S4Gru source. It may have gotten garbed somewhere along the way since Band 26 list not listed in any of the related 5G tables. I believe larger SCS directly improves transmission rates in Mbps, iirc.
I really liked another link you just posted about fine grain TDD. Hopefully this is used by Sprint in the initial 5G NR implementations as it might fix the band 41 upload issues some people have even moreso that having a 60MHz to 100Mhz carrier. https://medium.com/5g-nr/5g-nr-the-new-radio-interface-for-5g-2b769a59ea80
For those asking about qam. Does everyone for that cell sector use 256 qam? (in range of or phone support) If not your the only one going to see the 33% increase. So your old speedtest x 1.33. Examples 1Mbps to 1.33 Mbps or 5Mbps goes to 6.5 Mbps. Another would be 20 Mbps to 26 Mbps all within margin of error with speed test. That gain starts to go away fast as people use more data.
Where qam even massive mimo help is large events. Otherwise you have people on the fringe with weak signal weaking their effectiveness.
This is exactly what I am going to be doing. I have 2 galaxy S-7's both with deteriorating batteries. When the S-10 hits the shelves, surely the S-9 will go on sale somewhere. I either get a new S-9 at a good discounted price, or I get one on Swappa, or maybe buy the new "A" series Samsung phone. No S-10 for me. If the S-7's had HPUE, I would probably just get new batteries in the S-7's.
Frankly I think what cannibalizes their high end sales is picking up last year's model on Swappa or even used on a carrier's site.
We'll see how much of a value price point the S10e comes in at, but I'm guessing at least $700, so not much of a mid range value.
LG could really offer an attractive alternative, but they price their flagships just as high as Samsung, but then drop in value even more months down the road.