2) My example was clothing. I certainly think this makes sense as a setting in such an environment. Let you look and try. Directly sell most common sizes, transfer to online purchase for others. You can even have employees measure customers to get the right fits! Now you could even do the virtual tryons. This is very different than racks of clothes.

3) I think you forgot about stores like Sharper Image, Electronics Boutique, or Brookstone. Customers frequently would go into these stores to just see all the random gadgets and stuff. I can certainly remember going into Brookstone dozens of times yet not actually buying anything. Thing is, what these stores were good at was advertising products. But they were terrible at selling them because you could always find the same things somewhere else for cheaper, like Sears.

Like I've said, the value of many of the physical stores was not just in direct sales. That was a fine metric in the old days, but things changed and so did many other things. My original comment was a claim that a myopic view was applied, hyper focusing on the limited direct sales metric. But coke doesn't advertise to make you aware of coke nor do car companies advertise to make you aware of cars. They do things differently because their size and markets are different.

My point of a lemon market is that with the loss of ability to physically scrutinize products, you cannot tell the difference between a lemon or a peach. What I didn't say, is that this incentivizes more dark patterns like making returns difficult. Part of Amazon's quick adoption was free returns, making the downside of buying a lemon low, only costing you time. But the idea of tricking you into buying something, especially with a subscription, and making you live with the purchase sounds more like the strategy of an infomercial penis pill scam, not a blue chip business.