The modern failure points are actually cracked printed circuit boards (or lead-throughs) and cooked components because the mounting options aren't as flexible as they would appear (mounting over a fireplace is not an ideal location).
We can only blame ourselves for the heat issues. Manufacturers give ventilation requirements that everyone promptly ignores. Maybe you shouldn't jam that new receiver or BD player into a shelf space exactly the size of the case. Aside: most fireplaces are seldom used, other than as a selling point to market the house.
Everybody went ga-ga over that ultra thin tv at CES. What thermal compromises were needed to meet that spec. We just finished a thread where people were complaining because a certain plasma set had a cooling fan. These suckers need air to live, and we keep suffocating them.
That gets furthered by a certain product getting a reputation for running hot, or being noisy, so the manufacturer is forced to cut margins in order to address the rep. Everybody says Onkyos run hot, and people give anecdotal evidence that they are hot to the touch. After awhile, everyone believes it and it is then a "fact". A lot of it is bogus. First, I have measured my well ventilated Onkyo multiple times now and never found a temp that exceeded 105 deg F. Second, the statement "feels hot" is meaningless. For example, a metal case encourages heat transfer, both out of the unit and also has good heat conduction to your hand. A plastic case insulates the unit so heat stays inside. It also has poor conductivity to your skin. Thus the plastic case unit feels cooler to the touch, but in fact the interior temperatures are significantly higher.
Wave soldering is nowhere near as robust as wire wrapping.
First, when was the last time you saw something wire wrapped in production? Soldered lugs were less reliable if anything, because they were all individually soldered by line workers. I fixed more than a few cold solder joints in TVs in my youth.
Wave soldering is not less robust than through wiring, nor is it worse than point to point. If it was, MIL SPEC would not allow surface mount components. In my many years at HP/Agilent surface mount was never a reliability problem and we put those boxes out in mini subs in the bering sea, down volcanos, and even up in the ISS.
The problem is that ten years ago Sony decided to offshore manufacturing to southern Asia and went low bid on the contract manufacturers. These manufacturers seemed to think that they didn't need process or manufacturing engineers. Further, they turned the temps down to save energy. Result was poor bonding, and a reputation for bad reliability in wave soldering. (see reputation discussion above)
The average will go down because you simply can't get replacement parts and component level replacement usually means big money as each component is integrated with a bunch of others.
Are you trying to tell me that TV sets of the '50s-'70s were more reliable than today??? Perhaps you are old enough to recall the tube tester in every drug store and supermarket. And discrete components were no picnic either. They were more failure prone, and the current levels were high enough that any failure meant several other components blew in sympathy. Finally, that TV generally had all its capacitors dry up at about ten years, or the CRT went out , leading to a repair cost that exceeded the price of a new one, so no gain there either.
Larger components are expensive and often single sourced. They also obsolete quickly and there is no requirement for a service life in consumer electronics. That is a real problem. However, the components themselves are significantly more reliable
