The average turbo reported in userbenchmark is going to be significantly lower than the max rated boost clock. E.g. for the 3900X it's typically around 4-4.1 GHz, in contrast to the rated 4.6 GHz max boost. Given that the 3950X has 33% more cores but the same TDP as the 3900X, I'd expect the heavily threaded boost clocks to be lower than that. So the 16C TR could still end up having higher real world boost clocks than the 3950X even if these results are more or less valid.This looks very unlikely.
Given that TR can have 80-100% higher TDP compared to desktop, there is no way it should be at 700MHz lower clock than equivalent desktop CPU.
They would have to use total trash chiplets to clock lower than 1st generation zen cores.
On the other hand, the Ryzen 3000-series desktop processors launched a few months later in the year than their 1000 and 2000-series counterparts. The first generation Ryzen processors launched in March and April of 2017, with Threadripper coming several months later in August. The second generation came in April of 2018, with some Threadripper processors coming in August, and others in October. Third generation Ryzen didn't come until July, so if Threadripper again comes several months after the AM4 parts, we might potentially be looking at a release sometime around November.However, if AMD follows recent norms, it could release them next month. Looking back at Threadripper's history, we can recognize a fairly consistent pattern. AMD launched the first-and second-generation Threadripper processors in August 2017 and 2018, respectively. There's a good chance AMD could continue that trend.