They are very hot and will take a fair length of time to cool (not much cool mass to transfer their heat to and wave radiation will take time).
They are also very, very small. A boulder heated white hot will take a great deal of time to cool off, a spark spat from a fireplace, not so much. And let's not forget this is being pushed extremely hard away from the observer.
I don't know for certain but I'll bet that even a kilometer wide bell will not stop particles rapidly going beyond your ship's silhouette.
Actually I was postulating three kilometers in radius but honestly on reflection a few hundred meters would probably do it.
2) If you ever have to change course, you'll have hot particles still that are no longer shielded, further intensifying the problem.
Indeed, no arguments there. However this being space a savvy captain would simply burn for a while then let it cool off while he coasts, before changing the attitude of the ship to the new heading and burning again. Also keep in mind that the further away they are the more leeway they can allow in active course changes - ie the more particles they can let slip past the shroud.
3) A kilometer wide shield may in fact visually occlude things that will allow optical spotting.
At one million kilometers a six kilometer wide shiled will occlude about one eighth of the space that Venus does at its farthest from earth. That's the size of possibly actively cooled forward profile you're trying to detect, and if you're trying to detect it on a ship you're moving very fast while doing so. One million kilometers isn't all that far in space, depending on relative velocities ships could get to within fighting range in hours or days, perhaps even less. A mass driver could pump out a cold missile at extreme velocities to coast right next to a target before igniting from a million kilometers.
4) We already use synthetic arrays. Why does the article limit the array to 24m? I'd say 100m+ is feasible and that sizably increases detection distance.
As another poster mentioned, such equipment doesn't tend to respond well to g-force if mounted on a ship. But okay let's run with it.
If we assume tech progresses, we can have dispersed arrays (and should have given the possibility of attacks) using various satellites deployed in varying orbits around the system (including perhaps ever 60 degrees from your satellite in question plus some further out in the system). In fact, a coordinating system may be able to process data from every friendly sensor in the system. So both the assumptions about detection threshold/range and the assumptions about how many different perspectives at different angles might be available vastly changes the chance of your ship from sneaking in with a hot exhaust.
Ah now you're making assumptions about the environment, that the future consists of a tranquil hegemony of polities throughout the system, or that any putative conflict would even take place within such a system. Or that your fragile monitoring arrays won't themselves become targets in a co-ordinated attack, by their nature they need to remain relatively fixed.
Allow me to postulate a different but equally likely scenario.
Many different polities, on earth and in space, with varying relationships to one another, some cordial, some not so much. There may be corporate states involved, whatever. The interior of the system is well covered by surveillance networks but the exterior not so much, for political and physical reasons. Our space pirate heads off for the outer colonies, all well and good so far, then keeps coasting to a patchily covered location. They extend stealth shrouds, flip, and burn hard for the nearest cargo route, shutting off their engines in time to cool to background levels before they return to coverage.
The shrouds might also act to to deter active scanning. The ship itself goes into stealth mode, all non essential systems shut down, the crew are relegated to a sedentary lifestyle that many modern gamers would be familiar with. In extreme stealth mode they might even be refrigerated, which would allow the ship to cost for far longer time periods.
Once close enough to a trade lane, the pirate ship wakes up, lights up, and takes on the nearest freighter. Due to the distances involved help might be days or weeks coming. Alternately merchant ships could be heavily armed or have escorts, which leads to epic space battles. Taking the loot the pirate burns hard for the safety of the outer reaches where there's no surveillance looking in, or very little, and changes course in the emptiness, becoming once again invisible and annoymous.
I'm not saying it would be a situation which could last for long any more than the age of piracy did on earth, but it's definetely possible.
5) Real engineering means perfectly spherical ships that are thermally identical in all facings are pretty much just not going to happen.
How does that make it any less effective than all the rest of the imperfect engineering (which is all of it)?
Where you have a distributed detection network, as you will around any system of note (and around any fleet of note because they will tend to distribute satellites to extend their synthetic array and increase its sensitivity), you will find it very hard to sneak up with the ship the article mentions.
See above.
why can you drag a km wide heat shield and the other side not drag a km wide lightweight array?
Because one is literally tinfoil and the other is fragile optics and electronics?
My best guide on this is an instrument scientist I know who worked for NASA and who worked on thermal detection. His opinion, with his knowledge of current state of the art and what's likely in the near future, is that thermal stealth is going to be nearly impossible in short order. His opinion was real time processing was within grasp for full-sky in close to real-time within the next decade or two if we wanted to invest in it.
Okay well your anecdotal buddy should probably read this: http://www.businessinsider.com...
By the time we're out colonizing space, it'll be commonplace. The stealth can't keep up.
I say otherwise.
The hard fact here is that project rho is a scattered collection of out of context quotes, falsehoods and flat out glaring mistakes. Like in the laser weapons page, he says you don't need multiple laser turrets, you just need to channel your laser down a central tube and guide it with mirrors. Then a few paragraphs later he says you can't deflect lasers with mirrors because they'll burn out. In the x-ray lasers page he was caught out in saying there was no such thing as x-ray lenses - there are. And yet further down the same page he's still talking about the implications of a lack of x-ray lenses.
Have you seriously looked up the guy behind that? He ain't no scientist and there's precious little science on his page.