Their armor is pretty horrid actually. You need to remember that the IG is explicitly designed as a planetary expeditionary force. The local PDF is supposed to be the static garrison force, sufficient to handle the expected threat profile of a world. The IG is supposed to be the offensive and intervention force; deployed to take enemy held territory or defend territory under extreme threat.
What this means in practice is that the IG has to ship its armor between planets along with spare parts, ammo, fuel, and all of its other logistical needs. What that means is that IG armor should be designed to have the smallest logistics cost possible.
IG tanks should, given the IoM tech base, be all electric and using fusion for power. Use solar power to distill and then split water into hydrogen and oxygen, fuse the hydrogen for power, arm the standard tank with Las weapons and an EM rail gun. Take iron, either fusing up to it or extracted from the environment/sourced locally, shape it into rail gun rounds. Las weapons for anti-personnel work and for active defenses (shooting incoming rockets, mortars, and potentially even bolt rounds or lower flying aircraft out of the sky), the rail gun for heavier work. Ideally use a Las shot fired immediately before the rail gun to burn the atmosphere out of the way before the rail gun fires so that you can avoid air resistance. The sole logistics cost of such a tank becomes delivering the tank to the battlefield, everything else can easily be sourced locally and it is good enough for most IG needs.
IG artillery should be essentially the same, save using a more powerful railgun to deliver bigger and better shells. Anywhere with a breathable atmosphere should have all of the needed molecules in the air to turn into at least passable HE. Sure, you probably need an off world supply for more exotic ammunition but allowing organic local sourcing of your primary type of ammo is a huge reduction in logistical burden.
I don't think the Imperium has the technological base for that. Namely, they can't produce plasma reactors of the right scale to power tanks, and if they can, not nearly in the numbers sufficient to power even a tiny fraction of their tanks. Even most non-archeotech Astartes vehicles don't have those, even though rationally Astartes would be the higher priority users for those than IG. The smallest vehicles where plasma reactors appear as normal, i think, are Knights and heavy, long range, Navy grade starfighters like Fury. Which in turn would suggest that the vehicles to use those would have to be of similar size, and compete for industrial support with these mentioned vehicles - which are in turn known to be rather expensive, rare, and far more valuable on the battlefield than a simple Leman Russ, or even a Baneblade.
Also as far as logistical burden goes, spare parts, especially of kinds that need to be technologically advanced, even IRL are a far bigger logistical expense than consumables like fuel and ammunition. This goes double for the Imperium, where the production capacity for such parts is so insufficient that even the powerful and wealthy like Astartes chapters and Rogue Traders struggle to be sufficiently supplied with them already. Meanwhile, the average civilized world, of which there are many, may not be able to make plasma reactors, or even overhaul existing ones, but can make perfectly good promethium, multifuel engines, and battle cannon shells. Yeah, using those imposes additional logistical fleet costs, but Imperium does have transport ships with reasonable operation costs when considered per cargo mass/volume basis.
It's also worth noting that most of maintenance of basic IG vehicles is done by properly trained guardsmen, with one enginseer supervising them on higher organizational level, while such reliance on non-AdMech workers is unthinkable when it comes to a plasma reactor.
Then you have the lack of armory/forge/logistics ships at scale. One of the most common vessels in the IoM should be basically a mini-forge designed to produce everything that an IG army needs using locally sourced materials. Basically, one of these in system should be able to keep an IG army of, say, ten million fully equipped to standard for say a century using nothing but locally sourced materials that it can gather itself (i.e. asteroid mining, for example). The Imperium has the tech, they even build ships able to do just that occasionally, they just don't bother to do it at any kind of scale.
They can do it, and they have the ships (Goliath class factory ship). I guess its typical of the "too little, too late" problem, as these ships aren't considered very common, or very cheap, and this model (alongside with even more capable Ark Mechanicus) are mostly kept by AdMech for their own needs, and probably they whine about not having enough of them anyway.
And for planetary strike, take one of those armory ships, give it some rail guns, and have it mass produce proximity fuzed bolter rounds. Once an enemy battlefield or fortification is identified, just launch a few dozen million of those on a course calculated to deliver them to the needed grid square. Sure, it will do jack to shielded targets but most things aren't shielded and you can always just keep the iron rain crashing down around the shielded area for a few days/weeks/months/decades at a time. Counter battery fire on a void shield equipped vessel a few hundred thousand kilometers distant and constantly moving in random directions at variable accelerations is basically a pipe dream.
But then the IoM doesn't actually appear to have any dedicated planetary siege vessels. Its naval vessels are all primarily designed to engage one another in the void with any planetary strike missions either being very strategic (at a minimum removing an entire city) or requiring the vessel to be absurdly stupid (ah yes, we must bring the cruiser into atmosphere and park it in orbit over the battlefield for a full day before it can strike any of those puny ground-pounders with anything approaching precision). Take a cruiser sized hull and equip it to fire what are basically bolt rounds in the millions along with the needed computer support to handle relatively accurate targeting from trans-lunar orbit. Give it some kind of surveillance parasite craft that can be sent in ahead for mapping/targeting purposes, and then just unload on enemy grid squares. Drop a million bolter rounds into a square kilometer with them all fused appropriately and your problem is solved. Even a basic bolt round should be able to handle orbital re-entry, already has a rocket engine for terminal guidance, and can already be equipped with all of the needed fuses and payloads. Again, "That's a nice artillery park you got there, shame it just ceased to exist."
Between ship designs and RT mechanics, the closest thing to planetary siege vessels are Astartes vessels with bombardment cannons, which are noted for being precise enough to use for close battlefield support. Which makes some sense, as any significant enough planetary siege would be usually expected to have some number of Astartes vessels take part.
Even then, i think the Imperium is just simply underusing the orbital bombardment of normal vessels, or at least seems as such, because it doesn't make for much writing, and novel scenarios focus on situations when its not too feasible. Going by Rogue Trader books, an average vessel with average macrocannon battery can provide quite decent support on strategic scale, roughly equivalent to batteries of superheavy artillery with worse accuracy - but still perfectly sufficient to ruin a city, or decimate and rout a division no worse than a whole salvo of tactical nuclear weapons would.
Of course with average crew and targeting equipment, providing such support "danger close" to friendly units is not advised, better leave that to higher quality vessels, like aforementioned Astartes ships, or at least lance armed Navy ships with crack crews. For anything else, like an artillery park 30km from the frontline, i guess the writers are just too lazy to mention random IN escort taking potshots with its macrocannons, missing, accidentally hitting a random forest 20km away, trying again, evaporating a lake on the other side of the target, and finally catching half the division's dislocation area with the scatter on the third salvo.
Millions of bolts may be more of a overly expensive improvised weapon for this, as even normal bolts are noted to be somewhat expensive, nevermind guided bolts, which are rare and expensive enough that even high ranking Astartes can only sparingly use tiny bolter magazines worth of those.