GAE's main customer is Google itself as far as I know, so I don't think they have any problems with traction. They certainty aren't priced or have the features necessary to get much traction outside Google at any rate.
They're a premium service with very harsh restrictions and poor portability. I gave them a try for awhile, but ultimately they have too many drawbacks and have been plagued with terrible issues such as 20sec+ start up times on GAE/Java. It's just not a great option considering you can get the same performance as ~50 GAE instances with a single dedicated host at 2-4 order of magnitude less cost.
AppScale itself is pretty interesting though, but without the Google Bigtable backend, most of the forced restrictions don't make much sense and actually decrease performance and reliability compared to just developing for a linux VM directly.
"...with very harsh restrictions and poor portability"
You can say the same about web development, but the real question is if the benefits outweigh the drawbacks. As a one-man-band GAE is almost a necessity for me, allowing me to focus purely on code. As far as portability, I'm religious about wrapping any GAE-specific code/paradigms so I can hop onto EC2 or whatever pretty easily if I need to (actually did need to once). Certainly if I was leading a large team, we would explore more efficient and cost-effective options, but for me it's perfect.
There is also Google Compute Engine, which may mitigate the restrictions of GAE...haven't tried it myself yet though.
They should be teaching it to everyone and it should be the default - you start with app engine on any project and only choose something else if it can't be used.
And it's about as cheap as you could make something like that, If google itself finds it valuable, that's a argument in favor, not against, google having some of the best software developers around.
The only limitation currently is it doesn't have postgresql.
(I don't have anything to do with google. I just wish more people used it so more features could be added, and more open source software written for it, and I'd benefit from that).
I have not really had a problem with speed. I have a game on it that does 400ms client -> sever -> client.(authentication done in parallel thread) You can put Django on it. ITS FREE!
Curious about your argument here. Care to explain?
You then learn sql and build complex apps.
Learn about bigtable and build a viral app that scales to anything.
Learn golang and make your python code more efficient.
And you can do this all free to start off with.
The alternative - learn linux, web servers, db admin, security. Or php/mysql and pay shady hosters $20 a month and buy domain name.
If you're experienced and know all about servers etc, you can now ditch all that knowledge and outsource it to someone else and focus on higher value things.
So does Khan Academy. Ben Kamens, their lead developers, writes a lot of great App Engine tips on his blog: http://bjk5.com/post/54202245691/the-app-engine-way
[1]: And that statistic is more than a year old now (from 2012's Google I/O: http://gigaom.com/2012/06/28/google-app-engine-by-the-number...), it's a higher number now.