Those 100x 500 dmg missiles are 10 OP while the 1x 4000 dmg torpedo is only 2 OP and calls for a smaller slot size. Also, a Pilum launcher takes eleven and a half minutes to shoot off 100 missiles unless you have expanded racks, in which case it takes about eight and a half minutes. Most battles are shorter.
This is like comparing garbage trucks to tomatoes.
Missile spam is not a zero-sum proposition, like firing a torp generally is.
That 1X4000 torpedo misses / gets intercepted more often than not (practical hit rates are probably <30% when we include AI, not to mention torps never fired at all because the torp AI is so cautious), while about 20-40% of the Pilums will do damage, at a greater engagement range, with capabilities that that torpedo doesn't have, like finishing off multiple targets. It's a huge difference in real power
My take thus far: I tried various flavors of this out in Vacuum, and it's a pain in the neck to get it right, for a bunch of reasons, but mainly it was a performance issue for me, with the much bigger battle sizes, because dead missiles weren't getting removed from the battlefield, amongst other things. This is something I was going to fix, eventually, as it was causing some real problems; dead missiles in huge swarms (especially fast spammers like rocket racks) were a big problem, in terms of AI loading.
Ammo limits being off for all missiles is better, provided that feature is tweaked (say, a 1 second fade-out time after engines are out). It makes things simpler than the current system, which has suddenly gotten massively more complicated and has only improved gameplay marginally and erratically. Missiles have always had the pull problems associated with ammo-using weapons in general; if you can run them out, which was pretty easy with everything but Reapers / Pilums / Harpoons (sorta) then missiles were largely off the table and you could bring in your own missile shooters and have a huge advantage. Taking out the ammo problem largely fixes that issue.
But what we have here is, to be frank, utterly confusing and pretty arbitrary-feeling. For example, Annihilators seem like a perfect fit for endless ammo; they were always largely useless spammers with just a few squirrel cases; letting them spam all the time would make them a genuinely useful weapon as a no-flux spammer. But they have ammo limits. Pilums, which were already the best missiles for long-term pressure and were the hardest to pull completely, don't, which makes them really great, and massively better than an Annihilator.
A 5 OP Salamander completely
wrecks an Annihilator, but more importantly, if the battle goes over 2 minutes, it wrecks a Harpoon firing three at a time, which costs more.
Sure, that Harpoon might be really lethal- once- but it's not likely, vs. targets with some PD, or if fighters pull them, or whatever, whereas that Salamander is still reliable against just about any target, if for no other reason than its engine-killing effects, and will stay in the fight. I'm not saying that the Harpoon isn't a good alpha-strike; it is, in player hands (and occasionally in AI hands, when it launches 12 or more in a massive volley) but generally speaking, the Harpoon has lost a great deal of relative power and I'd use the Salamander every time, if I could just get my hands on 3-pack launchers. Even the single-tube version is a massive step up from an Annihilator that costs more, though.
I don't think the right answer is to nerf the ammo-less missiles or to buff the ammo-limited ones, though. They really are about as lethal as they should be. I'd just put ammo regen on the rest of them and then pick some times that make sense; probably 10 seconds for a Harpoon, 30+ seconds for Reapers, <1 second for Annihilators, 15 seconds for Atropos. Adjusting those times to fix balance is probably the best way to get back to an even keel on this and simplify play, which has become massively more complicated with these changes and the clip system and is probably a confusing thing for newbies right now.