Batteries will be a capital-intensive, low-margin, commodity business. Tesla is priced like a mega-growth industry disruptor. If they just turn into a battery factory, the shares will drop 90%. Tesla's upsides would be 1) a valuable brand (but others have good brands too), 2) autonomous driving technology (but Apple, Google, and a number of car manufacturers are also developing this), and 3) a first-mover advantage. But first movers often aren't the ones who make bank on a new industry. Google was maybe the 10th major search engine. Apple wasn't the first to make phones. Microsoft wasn't the first software company. Etc. The firms that make a ton of money are the ones that come along and transform it and do it better. It's much easier to improve an existing idea than create something from scratch. And anything Tesla does that is successful will be copied by others.
I think the difference with search and cell phones is that they weren't monolithic industries, at least not compared to auto manufacturing. In the open ocean (new industry), it's easier for companies to come out of nowhere and take the lead. Auto manufacturing/sales have much higher barriers to entry, both in terms of regulations and capital requirements, and are very well established.
EVs in the existing manufacturer/dealer network are also very disruptive because they have the potential to hit the repair side of dealer profit centers and the sales side of manufacturing profit centers. My sense is that manufacturers/dealers view competition with each other on EV production as a very possible lose-lose situation. They have to spend massive amounts on tooling, research, and capital, and the end result could be a product that hurts the profit centers of their customers (dealers) because of better reliability/less maintenance, and may also reduce sales down the road as battery energy density increases and outside companies start producing replacement battery packs.
I don't think we're at that point yet, given that most EVs have < 100 miles of range and a bunch are still built on platforms designed for ICEs, but once manufacturers hit a good combination of range, charging speed, and price, I could see battery replacements take a chunk out of new car sales. I think Tesla could hit that mark with the model 3 in the ~$30k-40k price range, and if they do, there's less incentive to purchase a new car compared to getting a new battery pack and driving your current EV for another decade or so.
People also don't realize yet that Powerwall is a failure. The batteries are just too expensive for it to be a viable product yet. Maybe in 5 or 15 years.
I wouldn't go that far. On one hand, people do overestimate the impact of new tech. On the other hand, the powerwall is the least expensive form of warrantied battery storage out there for someone who uses it daily, which is why there's so much excitement about it in the industry. It's not to the point where you can drop a bunch of them in your home along with some PV panels and pay less that what you would with a grid connection, but in some places (CA for instance) it's getting close. If I were building a home right now, I'd take a good hard look at self generation (demand side management + PV panels + a powerwall of few + a backup generator) before I paid for a grid hookup.
Edit - I'm not saying that Tesla won't fail, especially considering the challenges they're taking on, but there's a good chance they'll succeed, which is why there stock is where it is IMO. Even if I ignored that Paypal, SolarCity, SpaceX, and Tesla have already been pretty successful, Tesla's success has correlated pretty well with the volume of "Watch out Tesla!" articles I see in online media. If that trend holds, the 3 will do very well. ;)
https://teslamondo.com/2014/05/25/yipes-watch-out-tesla-move-over-tesla/