With modern optics, it is possible to make out any sizable object at rather long ranges visually. With thermal, even moreso. It's hard to hide against cosmic background temperature if you have any sort of engine, electronics, etc.
So, let's posit a best case missile:
a) launched from a railgun (no initial thermal heating, kept near 0K before launch, dark outer skin, decent launch velocity)
b) no manouvering until final attack phase (requires a stupidly non-manouvering target for a long time so your missile can close without having to use thrusters)
b) is pretty unlikely. The minute you know an enemy ship is out there (see my first observation - provided to me by a NASA instrument scientist of my acquaintance who worked on the NEAR mission), you'd start some form of manouvers just to eliminate the blind ballistic launch.
a) This may make the missile harder to spot. The chaff cloud would make it very easy to spot the cloud and the chaff cloud can't manouver like the missile (and the missile might hit the cloud to its detriment) so evade the cloud with your ship and you force the missile to manouver and thus reveal it and remove its cover.
The amount you'd have to manouver (how hard) is inversely proportional to engagement range. At longer ranges, a relatively mild set of course altering thrusts would rapidly ensure that no chaff cloud would come anywhere close and any missile would have to start manouvering to be in any position to hit you.
So your long range attack scenario is pretty unlikely unless your opponent for some reason has their sensors (passive) off.
Modern telescopes (not even future arrays) married to a decent computer can do fast sweeps of the sky and spot optical and thermal differences. Marry this to computers 20 years in the future and it'll be realtime (if today's take a few minutes). And once you've made the initial spot of an enemy ship at a distance, then you can focus your array and processing time drops further as you don't need a full sky sweep, so missile spotting becomes very likely even if they are trying to be within a few degrees of absolute zero (background thermal level) and not manouvering. If they manouver, you have them sighted.
If a missile has to manouver a lot to keep up with ships, it has to carry lots of reaction mass and burn it. Hotter signature, ejected hot trail, and more weapon mass devoted to propellant.
Covering areas of space with any sort of projectile saturation (or even a chaff cloud) is impossible at any range. The dispersion rate of the cloud and the manouvering rate of an enemy target at any range to speak of will be enough to render the amount of projectiles you'd have to lob as 'vast'. For ballistic projectiles, you'd have to predict the enemy location at time of intersect and set your scatter to attain your chosen volume at that particular point. Any enemy manouver to speak of ruins that option.
So, this sort of long ranged attack might work versus planets or stationary targets, but never against a military vessel that was aware of an enemy presence and most of the time they would be. Stealth in space is a fiction that's hard to justify given modern sensor tech and computer power and that available in the near future.
Instead, just simply closing under manouver and using carefully aimed energy weapons and perhaps close enough in some secondary batteries using railguns with high speed projectiles might work. Missiles could be used in swarms at short range.
For obfuscation, you can jam active sensors with EMPs or emissions, put modern passive thermal and opticals will largely render much of that EW useless. Passive targeting will be enough to hit cruisers and destroyers and so on.
It's funny - real space warfare with near future tech is horrible to game out. It is boring. Much of the action is closing and the nature of fights dictated by relative closure rates (a joust with one pass? an orbital gunfight at short range?). Stealth doesn't work. Planets and stations aren't defensible (the latter more defensible than the former but still hard to save). Fleets closing will be able to engage at medium distances hoping for a lucky hit and then at close ranger ranges looking to score significant damage with some statistical luck and being able to put enough joules on the enemy for enough time to get burn through or mess up their sensors arrays and so on. Fighters won't have much point except as added defenses for stations where they only have to manouver in self-defense and are just extra weapons platforms (probably drones not manned).
If you ever try to design semi-plausible ships in any of the more realistic space combat sims, you realize you have a lot of trade offs. Then you realize tactics are a bit limited as are the types of effective weapons. Then you realize the geometry of fights is everything and many engagements will take hours or days to unfold and boil down to whose gunnery tech works better because the sides will be using similar systems. Whoever brought the biggest fleet has the best chance of winning barring one side having a tech advantage in targeting or energy delivery downrange. Fuel (reaction mass) is a big factor in what you can do on the larger scale and in ship design. Energy constraints become an issue too.
It's not like the movies. There aren't 100 strategies nor boatloads of different technologies and ship designs. Time, distance, energy, fuel, and the ability to see a long way but only engage effectively over a much much shorter distance really shapes what can happen.