Both interlacing and tying and soldering were practices instituted when spokes were of far lesser quality than those available today (and before the importance of correct tensioning of spokes was fully understood).
When I began building wheels in the late 1960s, the choices in spokes available from the local bike store ran from Torrington and Union (galvanized, mildish steel) to Trois Etoiles stainless steel. Back then, in the days of 36-spoke wheels being standard, spoke breakage was not particularly frequent, but it wasn't a big surprise when it happened, either, occurring at something like the rate of flat tires. It's interesting that several changes in bike wheel technology took place at around the same time, in the early 1980s: stainless spokes started showing up at lower price points in a manufacturer's lineup of bike models; spoke counts in high-end wheels dropped from 36 to 32 and, later, to 28, 24, etc.; rims were manufactured with newer, stronger aluminum alloys; and Wheelsmith began selling spoke tension meters.
The availability of tension meters might have been the most significant change of all, where wheel building suddenly changed from an art to a science. Builders could still make blunders (e.g., a local self-anointed wheel-building guru used to build even the lightest wheels, using the most fragile rims, with non-butted 14-gauge spokes---i.e., what DT referred to in their literature as "tandem gauge"), but wheel longevity went way up.
Spoke breakage can still occur, of course, so it makes sense that interlacing of spokes continues to be standard practice. As for tying and soldering, whatever vanishingly small benefit is conferred (if any) is offset by the need to replace two spokes rather than one when a spoke breaks.
On the other hand, I doubt that anyone here would disagree with the statement that high-flange hubs have been conclusively proven to be superior to low-flange hubs. Or is it the other way around? I can never remember.