I've heard a lot of stories from my older colleagues as I'm too young myself to have been working in soviet times.
As there was shortage of pretty much everything and the military was priority, the best stuff was allocated for military use only. For example, transistors with KT prefix were commercial-use and 2T were military-only. The military-purposed ones were better (tighter tolerances), but they were not allowed to be used in commercial applications. Hobbyists were able to get some unofficially sometimes, but they definitely could not be used in manufacturing.
For commercial stuff (ie radios), the components were split in 3 categories: things you could use, things you had to get special permission to use (incl paperwork showing why any substitution is not possible) and the stuff you could never use. It was not building down to a price point like now - you were not allowed to use these components at all. The allowed components had very large tolerances, so the schematics had to be engineered to cope with both extremes. For example, the probably most common transistor, KT315/KT361, had h21e of 20-90 or 50-350 depending on variant. And taking apart whatever device manufactured after 1980-ish, there are very probably some KT315/KT361 inside. Your schematic solution had to handle the variance, whatever was thrown in ther during the manufacturing.
All of the companies were state-owned and governed during the soviet times (with some small exeptions at the very end). Some engineers, who worked in large and important engineering bureaus at the second half of eighties, told that these larger and most important bureaus were actually using western gear - tek, rohde etc, because these were much better. These were smuggled into the country, probably at the goverment level. Officially, this gear did not exist. Even if development was done on these, the type qualification and all official testing had to be done with soviet equivelents. Engineers of that times have told me, that they were going mad with passing the testing - the better western gear told that everything was passing, but the official tests with soviet gear failed... And they had to tweak the products to get the required results.
Copying western stuff was not only ok, but required. Old engineers have talked how some officials smuggled some western audio gear into the country and then brought to the engineering bureaus to reproduce. The only issue was that the western component base was so much better that they had to design the internals ground-up. Of course these clones were much lower spec, but still looked quite similar. For example, there are tales that the gov officials in moscow had smuggled some Sharp Optonica audio gear to moscow. The management of RET engineering bureau (Tallinn, Estonia) were called in to Moscow and asked why they do not make stuff like that. They were given orders to reproduce the gear. The gear sold as "Estonia 010" series were visually and functionally clones of Sharp Optonica SM5100, RP7100 etc. The Sharp amplifier used ic output stage, it had preamp and power amp in one box. Soviet engineers did not have ic amplifiers, they redesigned the device with discrete components. Their version was much larger in 2 separate boxes (separate preamp and power amp), both of the boxes were 5cm wider. But the overall design and functionality was quite well copied. The record player was unique because it had automatic track searching etc. The sound quality was quite good considering soviet stuff, but the reliability was just horrible. Designing that large functionality with high count of unreliable discrete components meant that the failure rate was very high - i've seen probably more dead Estonia 010 amps than working ones.
Digital and microprocessor design was also very different. Most of the 74 series etc had soviet equivalants (K155, K561 series etc). The microprocessors were either mask copies of western stuff or functionally equal redesigns. Some older colleagues tell that when microprocessors became available outside of military market, the first years they were only able to get their hands on broken ones - the boxes of processor trays came with errata sheets, that listed which instructions worked on that batch. You had to program the code on that batch of controllers with those instructions only. A year later, these sheets changed to listing the instructions that did not work.