Configuration: 2.4 / 5500M (8 GB) / 64 GB / 4 TB
Geekbench 5 multi-core 7177
Geekbench 4 multi-core 30453
Cinebench R20 3347
Blackmagic Disk Speed Test Write 2.8 GB/s, Read 2.7 GB/s
I have seen upload speeds as high as 35 MB/s (280 Mbit/s) while sending large files (a terabyte or so of photos) to iCloud. I'm not sure whether the WiFi network I'm on (a modern, high-performance campus network at a major university) can go any faster...
All the benchmarks we're seeing are very high numbers for a notebook, as they should be for the price and specifications of these machines.
To put it in perspective, a base iMac Pro has a Geekbench 5 score right around 8000, Geekbench 4 a little over 33000, and Cinebench right around 4000. The high-end i9-9900K iMac is very closely comparable to the base iMac Pro, with the 9900K reaching a somewhat faster Geekbench 4 score (36000+).
A top-end 12 core trash can Mac Pro is slightly slower than the MacBook Pro in Geekbench 4 and 5 (the browser is complex to search due to the presence of numerous Hackintoshes using MacPro6,1 system identifiers). There are some MacPro6,1 identified systems that are much faster, but listed as running 32 core Threadrippers, 18-core Intel HEDT chips, 28 core Xeons and the like... That's a Hackintosh!
The 28 core Xeons could conceivably be "real" 2019 Mac Pros using the 6,1 identifier instead of 7,1 to avoid drawing attention, but are probably also Hackintoshes. The problem is that there are Hackintoshes mixed in that don't have outlandish scores or configurations, so the 12 core trash can results will always be contaminated by some systems that aren't what they claim to be.
Apart from Hackintoshes, there are very few Macs faster than the 16" MBP. Any iMac Pro is (the base configuration marginally, higher-end CPUs by quite a bit), but the only other non-vaporware Mac that is faster is the hot-rod iMac (the i9-9900K upgrade on a current 27" Retina iMac).
Of course, the new Mac Pro will be faster when those results start hitting (prediction: the 8-core will be faster, but embarrassingly close to getting beat by a laptop, while upgraded CPUs will be quite a bit faster - with some of the high-core configurations being extremely fast - rising to the top of all Mac benchmarks, even displacing most of the fastest Hackintoshes).
Compared to Windows machines, it's tricky, since the vast majority of what's in the Geekbench database is Macs (searching "9980HK", about 80% of the machines are running MacOS and half of the machines running a Windows OS are actually Macs running Bootcamp). I looked through all the 9980HK results in the Geekbench 5 browser in hopes of finding a reasonable sample size for some Windows machines. There are only a few actual Windows laptops running the 9980HK that show up enough to get a somewhat significant sample (even going through all 120 pages of results).
By far the most common Windows result is the Dell XPS 15 7590/Precision 5540 twins. They average just a little slower than the Macs (just averaging in my head, the Dells average right around 7000 in Geekbench 5 multicore, while the Macs average around 7100-7150). One anomaly is that there are a few of those Dells that are significantly faster than any MacBook Pro. I suspect the Dells can be overclocked, even though they ship at stock clocks, and there are 10% overclocked machines in there?
There is another group of Dells that are clearly faster than the Macs - these are the big, thick workstations. The Dell Precision 7540 (15", 6 lbs) and Precision 7740 (17", 7 lbs) are both Geekbenching close to 8000, rarely exceeding that figure. I strongly suspect this is real, is not caused by overclocked machines, and has to do with cooling. The Mac doesn't throttle, but the inch-thick Dells with their big cooling grates on the sides can turbo higher. There were vanishingly few HP (only a couple) or Lenovo (none at all that I saw) big workstations in the database, but I suspect those would be the same.
The only other group of faster PCs were random 17" gaming laptops (not enough of any one brand or model to say anything about a brand), which definitely ran higher - but the few I looked up were close to 10 lbs... They certainly have more cooling, simply because of their size. I strongly suspect that some of them are also overclocked, either by default or by their owners? There were a few results over 8500... Apple makes a machine kind of like that (without the battery) - they call it an iMac (although the 21.5" iMac doesn't get the really fast processors - those are reserved for the 27"). Some of these monsters are actually heavier than a 21.5" iMac...
There were two other interesting groups of Core i9-9980HK machines, both slower than the MacBook Pro. One was the unusual designs - there were a fair number of ZenBook Pro Duos in the database, which averaged about 6600-6800. That's not a lot slower than a MacBook Pro that scores 7100, but it's definitely slower. There were a few HP ZBook Studio x360s, which are a "flip-around" design. Those were a small sample size, but seem to be quite a bit slower (average score 6000?).
The final group was Macs running Parallels, benchmarked in the guest Windows OS. Unlike Bootcamp, these were quite a bit slower than the Macs should have been natively. I was expecting a 10-15% penalty - it's more like 30-40%... Of course who knows what else was running on the Mac side - it's an easy error to benchmark a Mac in Parallels while the Mac is also running Photoshop...
Love that score number.