Here is 1 patent:
https://scholar.google.com/citations?view_op=view_citation&hl=en&user=6Q4QknsAAAAJ&sortby=pubdate&citation_for_view=6Q4QknsAAAAJ:Mojj43d5GZwCThe figures from fig1-6 depicts a few assembly steps. First components are placed on an adhesive sheet so they are all horizontally level. Then adhesive is applied on top and they are lowered in a negative template (machined or molded). The hole on top is to let excess adhesive escape. The adhesive around the parts are cured, and the sheet on the bottom is removed. How the adhesive finds it way around the part without leaving air bubbles, instead of all going through the vent hole is a mystery to me.
This patent refers to the final assembly in which an anisotropic conductor is used:
https://patentimages.storage.googleapis.com/fc/dd/f9/f31bfd82598b7a/US20090017264A1.pdfIt basically looks like some kind of a zebra strip material. I suppose the material has very small vertical conductive pillars that are distributed semi-randomly so that any top point can be find conduction to the bottom part of the sheet, but due to the small structure size does not short to nearby pins for a small pitch layouts.
I do wonder though.. sure this may be solderless, but it looks like it removes all options of heatsinking for a kind of component. We don't know the thermal conductivity of the anisotropic conductor to the PCB, but I expect it to be fairly poor. The component is molded in, so it would need special provisions to add a heatsink afterwards. Since all electronics consume some kind of power, especially power supplies, I would see that as a major concern. Ironically the thermal stress is one it tries to solve.
The website talks about the frame being an integral heatsink to the design. But personally I think: compared to a conventional heatsink, you now add a potentially relatively thick substrate and additional adhesive that needs to correct for machining tolerances of a whole PCB (instead of only the tolerance in 1 part). I don't think those 2 combined can be "better" than what we now have, as it was not it's initial goal. It would probably be very good in distributing the heat around the whole PCB though.
Second I wonder about the structural rigidity. We still like through-hole mounted connectors because they can be much stronger than SMT. Sure the component frame may provide extra stability for the connector at the point of largest leverage, perhaps it can also add as a very good mechanical shock conductor and introduce new problems. It may not be solder related, but with respects to bad contacts, metal fatigue in braided wires, etc.
And then clearly this concept isn't going to work for all components. Straight up connectors can't be covered by the frame, but how would it be secured in place? Right angle connectors only? Well, FFC connectors also require a lever operation so need to be partially free. Etc.
I suppose it can work out, but I'm not convinced that removing soldering is going to suddenly make electronics a lot more reliable.