So, it's ages since I griped about Deathwatch! But I had some more thoughts due to working through our latest podcast (I foolishly and accidentally posted the post-podcast post way ahead of schedule - now taken down, with apologies).
One of the things that struck me is that the skill mechanics in Deathwatch further exacerbate some of the issues I have discussed by dint of where the designers choose to build in bonuses and penalties.
Let's use Brother Iacomo and his heavy bolter as our first example. He begins with a BS of about 40, and we can reasonably assume that he will rapidly start buying up BS enhancements to at least 50, which is easily achievable as a starting character. Many devastators will take the Immovable Warrior ability for an additional +10 whenever they are in cover. If possible, Iacomo will prepare for the shot by spending a full action aiming for a +20 bonus on his next attack roll. Firing on full auto, which is almost inevitable, Iacomo can gain a +20 bonus to hit; semi-auto will offer +10 to hit. If the target is larger than human - such as a xenos monstrosity or any significant group of weaker targets modelled as a Horde - there will be a modifier ranging from +10 to +30. Should Iacomo be facing an onrushing horde at close range, another +10 to +30 is available. A variety of targeting sights also offer +10 bonuses in specific situations. A signum and signum link can be purchased for an additional +5 or (if he's lucky) +10 bonus.
Firing a heavy weapon unbraced will incur a -30 penalty, but this can be overcome by spending a half action bracing, buying the Bulging Biceps talent, or purchasing a Deathwatch suspensor. Given that a Marine with a heavy weapon is presumably intending to fire it, few Marines who choose to focus on heavy weapons will not take one of the latter two options. Obscuring weather conditions, and certain kinds of atmospheric pollution, may also impose a penalty of up to -20, but typically a Marine can ignore such modifiers due to heightened senses and the technology of the power armour.
In short, Iacomo can, under optimal conditions, increase his effective Ballistic Skill from 50 to 180, which more than triples his skill and makes failure essentially impossible. More commonly, he will benefit from a subset of the available modifiers, taking him to around 100. Failure is extremely unlikely in these circumstances. If Iacomo chooses to further increase his Ballistic Skill, he will do even better - and given that shooting things is his primary job in the team, this is a no-brainer.
Similar modifiers are available for melee and psychic attacks, although there are fewer than for ranged attacks.
At this stage, let me flag that these modifiers are pretty important, because skills in Deathwatch are kept artificially low by the insistence of making percentile statlines look like the statlines from the d6-based tactical wargame that uses cross-compared tables. Since combat is also lethal, characters really need those bonuses if they're to survive and thrive. Equipping and planning properly makes all the difference between a walkover and a TPK. This is perhaps a unique feature of combat skills, as most others do not directly and regularly deal with potential death.
Let's now grab Brother Conscientious and get him to apply some of his hypno-learning. Brother Conscientious has made sure to increase his Intelligence and Fellowship to 50, so he's presumably a Space Wolf Rune-Priest or some such. Faced with a tricky technical task, Brother Conscientious prepared himself by spending a full action concentrating, gaining a +0 bonus because no such action exists. No starting abilities offer any bonuses to such skills. Conscientious can spend some XP to gain a +10 bonus to a specific skill, but only a small number of skills using Chapter-specific advances; most of these are rarely used, with the Tactics and Lore sets being most useful. Another talent offers a +10 bonus to Inquiry, Tech Use and Common Knowledge, but only if he's a techmarine. He can spend yet more XP to gain a +10 bonus to Intimidate and Interrogation, at the cost of dropping his Fellowship the rest of the time. One or two pieces of equipment offer a +10 bonus to a specific skill at significant cost: it costs 15 requisition to gain a +10 bonus to Tech-Use, a skill used about once per session, whereas most weapon sights cost 10 requisition and can be used for many shots in the average game.
Difficulty is more complicated. The rulebook helpfully provides only a mere handful of specific examples of difficulties for non-combat situations, and so it's very hard to tell what modifiers might be appropriate. In the sample mission, convincing an injured tech-priest that you are allies imposes a -20 penalty on Fellowship, even though this appears to be exactly what Fellowship is for, while doing so using the very specific Adeptus Mechanicus skills is at a +0 rather than actually offering a bonus. Tracking a character is at a penalty due to poor light and difficult ground; analysing some blood if anyone miraculously has the Medicae skills is at a -10 even though it seems a pretty basic task; extracting a small amount of data of limited use imposes a ridiculous -20 penalty on Tech-Use, which will be no more than 40 for most characters in the first place and perhaps 50 for a dedicated Techmarine. Persuading an annoying civilian to leave you alone offers a +20 to social skills, while obtaining a detailed security report from a terminal is at a +10 despite seeming more difficult than the previous Tech-Use skill. In short, I don't really have any idea what the difficulty levels are intended to do, unless it's to basically ensure things go more or less as the GM expected.
Thus, Conscientious seems to be able to get maybe a +10 bonus to skill if it's Chapter-relevant or one of the few skills with a specific tool that he bought, and honestly can't realistically expect more than a +30 bonus from circumstances unless he's on a library world with a team of dedicated sages who happen to specialise in the specific task he wants to do (not that kind of task, but that exact thing). More reasonably, favourable circumstances might occasionally offer a +10 bonus, and more often he will actually be acting at a penalty. In my experience of scenarios so far, penalties are extremely common (including a full-blown -60 at one point) and bonuses nearly non-existent.
Compare and contrast
Essentially, the point I'm highlighting is that the skill mechanics reinforce combat ability by offering a range of positive modifiers to combat actions, thus allowing a skill of 40-50 to translate into a very high chance of success if the character acts sensibly. In contrast, there is typically nothing a character can to do boost their chances of success at non-combat actions, leaving them with the baseline skill, or frequently even less. This combines with a number of other factors to leave these skills languishing in failure.
One is, for most Deathwatch characters, the very great expense of getting good at skills that are not exactly what the designers planned for your speciality, which can easily be triple the cost of enhancing the skills they did plan for.
Another is that in many cases it's flat-out impossible to raise these skills, which don't appear on most experience charts other than in basic form, if they show up at all. Many rely on Intelligence, Fellowship and other lesser-used stats, which often tend to be lower than other stats (because you tend to build a character based on your good stats, and a high-Int character is likely to become a Techmarine or Apothecary).
The third is that most such skills are rolled only occasionally, and combined with low success chances, this leads to a rather swingy experience. Because they're occasional, and have few positive modifiers, it's not really worth spending lots of XP to boost them; because they seem swingy, players don't build playstyles around using them when shooting things is much more reliable.
The range of modifiers available for Deathwatch play thus helps to direct players away from making extensive use of non-combat skills. No matter how much effort characters may be willing to spend, it is typically very difficult to get a situation where these skills are reliable, and as such it is inadvisable to either build a character around them, or make them a lynchpin strategy in handling a specific mission.
The design counterpoint I can take from this is that where modifiers are readily available - specifically, positive ones - this should generally encourage players to make use of skills even if they are relatively weak, since they can use equipment, tactics or planning to overcome these disadvantages. Similarly, a system where it is relatively straightforward to claim modifiers, providing you're willing to make the effort (at chargen, shopping or in-game planning) should help to encourage relatively broad skillsets, because pumping points into one specific skill should be less essential.