General notes from a Skype discussion with Marco last night:
General Dice Rolling: The proposed mechanic seems fine to me. We had to butt heads for a while over magnitudes of failure, though.
----Suppose you're making a Willpower test and you need a 65. If you roll a 12, you pass by 12 points. Simple enough.
----If you roll a 98, you fail by (98-65) = 33, which is what it is at present.
----So for a pass, you pass by how many points over zero you rolled; for a failure, it's how many points over your target number.
Actions and Reactions: This is a bit trickier. Early thoughts included defining what each type of "stored action" might be (so giving examples of how Sacrificed Actions, Wary Actions and Ready States work), and having the upper limit for stored actions be dictated by your Speed stat. Again, early days.
Risky Actions: Turning currently Risky Actions into Hazardous Actions seems like a decent idea as it eliminates the messy higher-Speed-means-greater-risk-of-failing-Risky-Actions thing we have at present, although we need to define what constitutes a Hazardous Action and how one can fluff it up. Providing examples would help to illustrate the point, but it does currently revolve around individual actions rather than the action roll at the start of the turn.
----How to define fluffing up a Hazardous Action? In the RIA, it's currently "if the units die comes up as a 5", though if it's percentile I'm more partial to "if you roll a double".
Movement: I had nothing meaningful to add here, but the jury's still out on whether getting rid of Sprinting is a good idea or not. I'm tempted to keep it, but impose a hefty penalty on Awareness checks and such like, as your focus is on pegging it.
Damage and Injury: I had a few things to say here.
----The Injury table needs some serious cleaning up. This is abundantly obvious. Taking the Abdomen chart as an example, we currently have a lot of "see the previous level, which redirects you to the previous level". That's a lot of messy bookkeeping that could and should be simplified and streamlined with discrete entries that don't refer you back to an earlier point.
----A suggestion I had was to scrap the current "one point into the next injury level forces that level of injury" mechanic, and introduce "floating injury points". Say you have a BIV of 6, and you're hit in the abdomen for 14 points of damage: that becomes two full levels (14/6 = 2.something), and then the remaining damage becomes "floating injury points" that can be transferred to other locations on the next wounding hit. So a subsequent hit to the chest for 4 points of damage might just add to the floating injury total, but because we already have two points saved up, that 4 becomes a 6 and that's enough to cause a Light Injury.
----I realise that the above means more bookkeeping and generally tougher characters, but sorting out the injury tables should balance that out, and it also gets rid of Four Slaps To The Head.
Close Combat: More actions and action types would be useful, but there's a danger of overcomplicating things if we go too crazy.
----Movement actions would help immensely.
----I took a hint from a couple of other games I used to play, where weapons were subdivided into such categories as "slashing", "stabbing", and "blunt". This has two implications.
----Firstly, we could incorporate "slash", "bash" and "stab" actions, each with their own additional effects. Stabbing actions, for example, could increase the likelihood of scoring a Critical Hit.
----Secondly, some weapons could be naturally suited to certain types of attack action and provide some measure of benefit -- a hammer would be a poor choice for stabbing someone, but ideal for inflicting blunt trauma and causing Knockback.
----I will freely admit that this needs more thought, as I came up with that at around 11pm.

----This branched off into ideas for a Close Combat RIA, or creating melee weapons where things like power swords and shock mauls are "upgraded" versions of normal swords and hammers. Needs more work.
Shooting: This needs more discussion, but basically, the current range modifier chart is a complete mess.
----I proposed a short/medium/long/extended range chart, so that each range band has discrete sub-ranges. That would be paired up with a single fixed modifier for each range; it wouldn't matter whether your weapon has Range A or Range F, firing at long range would still have the same fixed penalty, but what Range A might call long range, Range F might call medium or short.
----That ran into a problem where half an inch one way or the other might determine whether something gets slapped with a light modifier or a significantly heavier one.
----Cutting down the number of range categories (not range bands; I'm talking about 0-5, 6-10 et cetera) would be a compromise, making the table less messy without sacrificing too much granularity.
----What we do with Semi and Full Auto then depends on what we do with the range modifier chart.
----Aiming could be reined in to +10 rather than +20. This brings it more in line with Concentration and stops the silliness of a character's BS going well into triple figures*.
*At the most recent INQvitational, Sergeant Visstra managed to stack up no fewer than seven aim actions and ended up with an effective BS of 203. The DMR he was carrying made a very big mess of his unfortunate target's head. We decided that this made sense given the circumstances, but was otherwise a bit silly in practice.Psychic Powers: Different systems can at the very least be used as starting points. This needs more thought.
----The "power dice" suggestion would only really make sense if we ported in Dark Heresy 1E's system (which IIRC was itself ported from WFRP 2E?). I'm personally more partial to Willpower tests than power dice, as that preserves a measure of the existing system, but then again, we can always playtest different options.