no

Not only that, but the bitcoin developers have tried to distance themselves from mining entirely, not adopting the efficient forms of the high performance mining protocols that have evolved around high mining hash rate speeds, instead dissociating the work of mining to others (indirectly pools). While this may seem crazy, it is also indirectly accepting that it is virtually impossible for solo mining to make any sort of reasonable sense any more. I lament this fact, for I still think that all miners should have solo mining as their final backup should all pools fail, as a way to guarantee the bitcoin network remains secure, if not necessarily smoothly/consistently profitable for its miners.

Alas all of this is wrong.

All you need to do is add the following to your command: --gpu-platform 1

The cgminer documentation has extensive information on how to tune a 7970 for scrypt in the SCRYPT-README.  I spent ages writing documentation, please read it. It even states there that a rate of 745kh was reached, and I mined at that constantly for a few weeks before giving up on LTC.

79x0 devices fall into 2 categories: Those that like ultra high thread concurrencies (>20k) and ONE gpu thread, and these work well at intensity 19 or 20, and those that like much lower thread concurrencies (8192) and TWO gpu threads, and ANY intensity over 13 is going to cause hardware errors or fail to start. Generally speaking, all the overclock models (Ghz edition etc.) prefer the 2 threads while the reference models prefer the one thread. 7950 have no such reference model so it's a mystery which belongs to which category.

That is definitely far more than you should be getting. There is a good chance you're submitting heaps of duplicates which may also be a different form of instability that the auto mode can't check for. Try setting a lower maximum speed if you're using auto mode because clearly you're not doing 82GH of useful work.

If your usb hub is a usb2 hub and not a usb3 hub you will be far better off moving to the latest version of cgminer instead. The only reason to stick with the old version of cgminer is if you have multiple usb1.1 devices on a usb3 hub for there remains an instability there that we've been unable to debug so far.

cgminer no longer uses ttyUSB0 so you're changing permissions on an interface cgminer doesn't use. See the relevant readme instead (FPGA-README or ASIC-README).

Sorry, couldn't resist. You didn't deserve that comeback 

I used to get 1200W of heat from my GPU rig which mined 2.7GH of bitcoin. Now I get 1200W of heat from my ASICs mining 200GH of bitcoin.

This is all opinion based...

I estimate based on finding 5 blocks on average per difficulty change to avoid the variance of missing blocks before a diff rise in a way that can never be recovered. That's 5 out of 2016 blocks so you need to be >= .25% of the total network hashrate. At (current) difficulty 26 million, you need to generate 5 * 26 million shares in less than 2 weeks, or 26 million shares every 403 blocks which is ~4030 mins (in reality less time because diff is rising). That's ~6500 shares per minute. You get 1 share per 71.6MH per minute, so you'll need ~6500 * 72 = 470GH. One BFL SC minirig therefore is about the minimum requirement (this week).

Not only do you need a ludicrous amount of hashrate to reach that milestone, it gets increasingly difficult to set up an efficient solo operation as the network hashrate increases. Even the most hardcore miners gave up solo mining for any extended period for the risks are too great with such a large investment.

Summary: Pool mine.

No such thing as "appropriate" as it's just another hardware error. If your hashrate is okay, it's okay. You can read a few pages ago about how I chose the "appropriate" hardware error target rate of <2%.

You're going to have to give a lot more information then before we can help you.

Have other pools lost interest in implementing reconnect support?

You found it then. At some stage the avalon disconnects due to some USB issue and cgminer automatically hotplugs it. Since it's newly hotplugged it comes up as a new device (ava1 instead of ava0).

Precisely 14 microseconds after the first one of these is actually delivered, the price of the Asicminer product will magically drop to match it in price/performance.

Your password is confusing the conf file reader with the asterisk (*) there. Change your password to something without an asterisk or learn how to escape the asterisk in a json compatible way.

Cgminer submits stale shares unless you use --no-submit-stale. Some pools accept stale shares for some period after the block has changed to fake it looking like their stale rate is lower than other pools. Overall since they accept the same proportion of stales from everyone, no one actually gets paid more, it's just marketing.

It's got nowhere to go. It prepares work and then no device is working to take it so it just discards it.

Common reasons:
Wifi kernel problem
Overdoing the overclocking
Pool failure and the cgminer-monitor watchdog is trigger happy and kills cgminer when all it's doing is waiting for a pool to come back online.
FPGA failure in the avalon.

For maximum stability while I'm away, I limited my batch 2 overclock to 333 and it's been up for a week. At 350 it would restart cgminer every 12 hours and eventually after a few days would restart the entire router.

Yes