I worked in the defense industry for 30 years then taught college for electrical engineering, with lab courses that involved soldering, which we taught. As I remember the details, mechanical integrity was important, Tektronix not discounted, followed by cleanliness (I taught in soldering "Cleanliness is next to godliness"). Solder has one primary duty, which I havent seen mentioned explicitly, only implied: It keeps Oxygen from attacking and corroding the joint. A crimped connection will work for a while, but after 10-20-50 years, no! Hence the ARC-5 stories (I have one...) and Glenn's KWM-2. Ma Bell didn't want to solder subscriber numbers in, and so perfected the requirements for wire wrapping, also used in the early computer and microprocessor world to great effect. There, the tight wrap and the square post cutting into the wrap wire sufficed to keep Oxygen at bay. Yes, solder MAY improve conductivity, AND it MAY provide the necessary MINIMUM strength, but it real use was a conductive sealing agent, easily applied/removed.
Our technicians at Westinghouse Defense had to pass a 40 hour, 1 work week soldering school with a STRINGENT teacher (no tardiness, door locked at 8 AM!) before their job was secure! I still have the course books, as an engineer and a radio nut, I was interested. The group I led built widgets for our systems (a clever business ploy) and all of our work was microscope inspected (boy, joints sure look different under a B&L stereoscope!). I built many 8K memory cards for my S-100 buss system (using 2102's salvaged, Augat sockets of course), microscope soldered, 1000's of perfect connections! Looked like production reflow work!
Just remembering... Regards, Jeff Kruth