mmmm
If they are using a directional gain antenna would that not exceeed the ERP limit, so is it a real record ?
’ a bazooka dipole antenna, consisting of two hand made 8.4 cm (3.3 inch) elements connected to a TTGO node with LoRa module. ’
yes… its a REAL record imho 
(and probably hard to beat without ,for example, quad yagi auto tracking super high gain groundstations)
But the first ‘record’ claimed was;
The record was set by the PaPe I, experiment built by Jose Manuel Cuesta with that goal in mind.
The experiment used three 3D-printed directional Moxon Antennas, spaces 120º apart, and a reaction wheel for attitude control to prevent the payload from spinning underneath the balloon taking it aloft.
That ‘record’ was then broken shortly afterwards.
pure evil imho … and its growing 
Unmanned saildrone is first to circumnavigate Antarctica
TheThingsNode-2.0.8 Arduino Library released
Trying to understand the rationale for this. I am not sure I see the difference between the -078 module + well-written LoRaWAN API as we have now and the -104. What does the latter offer?
Can you (or anyone) summarize the announcement marketing-speak into concrete value-added statements? I mean, what is the purpose of this? It seems to be the -078 with SEMTECH firmware. Who would want that and for what purpose?
That seems to be about the size of it…with easy access to back end cloudy stuff by sounds of it. Many (larger/corp) customers will take comfort from the fact that the sw is from the lead LoRa co vs (admitedly potentially better/improved featureset/potentially app optimised! 
) version from a small design house or consultancy…
It’s an odd duck. It seems the idea is to use the module as a slave to an MCU host, since they mention one can communicate with it via UART. I just can’t imagine a large demand for such a device.
Large/corp customers might not like the “pig (or duck) in a poke” aspect…