40% sounds pretty close.  BTC Guild + Slush + Eligius alone aren't far off from it.  There certainly aren't any LARGE pools that are hiding merged mining as far as I can tell.  You can check yourself though by looking at blocks solved by "non-merged" pools.  You can't hide the fact that you're merged mining if you post which blocks your pool is finding.

I do not have any plans to support I0C merged mining.  The fact that anybody has the slightest interest in I0C at this point is baffling.  It's got nothing to offer.  The only reason I implemented NMC merged mining was that NMC serves a purpose outside of being traded for BTC (even if most people don't utilize that purpose).  Nothing like that can be said of any other altcoin.  Piggybacking chains onto BTC when they are technically a competing chain is just stupid.

BTC Guild will be adding support for multisignature transactions in the next couple of weeks.

The rejects are mostly duplicate shares, and slightly higher (not much) than average stales.  Overall they're still measuring above your advertised speed, although some improvements to stale share recognition could bring them up to par with GPU miners like cgminer.

Starting to come back online.  If the attack retargets the pool it may end up getting turned off again while we wait them out.

Server is still under attack, the second I boot up the server I can see the data/packet rates skyrocket into a completely unreasonable number.  Takes less than 10 seconds from the time the server is online before it is essentially unreachable.

If the server can't recover in the next 12 hours I'll try to get the server moved to a different IP address so automatic payouts can be processed at least.

Was hoping after the huge price drops and overall network drops these kind of attacks would go away, but obviously not.  Had a good run of stability ever since the PPS switch, but all we can do in the face of a DDoS is wait.

If you don't have backup miners, I'd recommend pointing to a smaller pool (I won't name any since most smaller pools can't handle a 500+ GH/s surge).  If past attacks are any reference, it probably will be at least 12 hours before the kiddies get bored and point their zombie army somewhere else.

Currently getting hit by a DDoS again, trying to come up with a solution right now (moving IPs or waiting it out).

On the phone with them right now trying to get this worked out ASAP.

Routing issues at the ISP side, trying to get info on ETA for being fixed (though odds are it'll resolve before I get an ETA).

Reject rate is higher than GPUs (Hard to compete with < 0.1%), but even so the speed I'm measuring is valid shares only.  After the first 17 hours, Median Speed is 200.43, Mean is 196.99 [using the 1 hour averages provided on performance charts with BTC Guild].  Peak was 240 MH/s over a 1 hour period, low was 169 MH/s over a 1 hour period.

I'm compiling a lot of data on these, hopefully I'll be turning on a small cluster (11 chips) next week and be able to provide people with a good sample of speed readings, averages, power consumption, rejection rates, etc.

Unfortunately, your first point is accurate.  The sheer number of infected machines is massive.  As much as I would like Bitcoin to obtain widespread adoption among the masses, the threat of malware at the moment makes it impossible due to the number of computer illiterate people.  While the average user is slowly becoming more security aware, there is still an alarming number of people that will type the same password on every website, provide details to every phishing email, etc.

Of course, the obvious solution to that is Bitcoin Banking, so the responsibility of securing their money is removed from individuals, just like most other currencies.  The only problem with Bitcoin Banking is that it starts creating the risk of undermining other selling points of Bitcoin.

I've added an entry on the How to Connect page for users of the Ztex FPGAs, and a link to the download page for the latest BTC Miner for those FPGAs in that section.  Just got my first batch and so far they've been mining flawlessly for the last 18 hours, with better than advertised speeds.

Anybody interested in a long term BTC mining presence should definitely check them out if they're in an area like me where the power costs prohibit GPU mining at the current price:difficulty ratios.  You really can't beat 190-210 MH/s for less than 10 watts of power .

Got my FPGAs earlier this week and finally got my 12V adapter for individual testing.  So far everything is working great, getting higher speeds than advertised!

UPDATE:  Added a link to my performance graph thus far.  Speed data points are using 2 hour average speeds on VALID SHARES ONLY.  https://i.imgur.com/xj6xv.png

The variance shown is within expected norms.  Peak is just over 220, low is just under 180, which means it has about a +/- 10% spread over the mean which is just under 200.  This is the same variance you expect from any GPU miner when taking slices of 2 hours for your averages.  I'll update again in a few days once I have enough data points to use 4 hour averages, which eliminate most of the peaks/valleys seen so far.

Hopping is pretty much just plugging pool stats into a formula, with maybe a little personal bias thrown in based on either past experiences or reliability between the pool and a particular user.  So you would expect hoppers to travel in groups, even if they're not related.

Low information density is an understatement.  Even the classic design had a lot of "empty space" forcing the site to be much larger than what it needed.

The idea of saying you're releasing a "low resolution" version later is laughable.  Right now the site doesn't even fit the data in 1920x1200 resolution vertically.  There is no reason for the amount of padding on every single element of the new design.  I thought the entire countdown was pretty unprofessional to begin with for a FINANCIAL WEBSITE.  Then this new layout comes out and as others have said, it has the look and feel of a blog more than anything business related.

My withdrawals from Mt.Gox hit Dwolla within a few hours at most.  I almost always have the cash in my bank account the next day [I setup Dwolla to auto transfer incoming funds into my bank account].  My last withdrawal was Friday and it hit my bank account Monday.

Not directly answering your question, but I think this might be relevant to a lot of similar question.  Additionally, please do not take this post as an attack on your question/you as a person, it's just something I've wanted to get off my chest as I lurk the threads for the past month.

<Rant>
Your brain lies to you.  All day.  Every day.  It makes you see/notice things that aren't there because of "patterns" that may or may not exist.  The human brain is determined to create order out of chaos in order to line independent events up.  There's some great lectures out there, including one by Richard Dawkins, that talk about the brain's predisposition to applying 'models' to information and sensory inputs.  In the same way an optical illusion can make you see things that aren't there, you can see patterns in a few points of data, often times overlooking contradictory data simply to help the model fit in your head.

Now somebody could come in the very next post and demonstrate a trend of the pattern you recognized, pointing out it has been been consistent for the past 10, 20, 50, or 100 iterations.  It may actually be a problem and your post brings it to somebody's attention.  But hopefully somebody reads this and the next time they think they notice a pattern, they take some extra time to think about it and do some research on their own for a significant data sample before talking out patterns.
</Rant>

I would just like to stress what was said earlier:  MAKE A NEW WORKER WITH A DIFFERENT PASSWORD!  Otherwise you're sending your login and password in plaintext.  Nobody should ever be using a worker with a password even resembling their actual account password.

That is why even I don't keep track of it.  A huge reason for the switch to PPS was to reduce the mental fatigue of monitoring round lengths and freaking out everytime we had a long round.  I monitor profit/loss (which is just luck in the end) on a much broader scale now.  I have no control or influence on it, so obsessing over that kind of stat is just a waste of energy.