In general I really like the abstract Influence-based requisition system, especially between missions.

I have problems with it though, when trying to use it INSIDE a mission.

Let's imagine the acolytes investigate in the underhive, lost all their weapons to a gang and now visit the local "store" to get at least some unreliable weapons at all to plan their righteous retaliation.

The store has some basic weapons available, which the acolytes would like to buy.

How do they buy them ? With Influence ? Does not really feel right inside a mission, if it is about buying something regular in a shop.

Furthermore this can lead to ridiculous situation, where a failed Influence throw to aquire a stub revolver fails, but the influence throw after this one in the same shop from the same acolyte can requisition a melta (success in second throw).

And in both terms he doesnt know before if he has enough money to get it...weird (needs to cram his pockets before).

Why not introduce an abstract "Money" good, which can be requisitioned during the mission by the acolytes, and can be used (and reduced) in an abstract way during the mission:

Especially if you want to keep subtlety.

Money categories and modificator on requisition:

1: no test needed (enough for a train ticket home after the mission is finished)

2: +30 (enough for a stub revolver)

3: +20

4: +10

5: +/-0

6: -10

7: -20

8: -30

9: -40

10: -50 (excessive means)

You can only requisition money once between missions (besides special situations granted by the GM), but once per acolyte.

The important thing would be: still keep it abstract and handle it as GM how much which shopping trip lowers your active money category.

You buy too much compared to your category --> your category goes down.

If acolytes buy something together, categories could be added together at the GM's discretion (GM determines the final level after addition, no 2+2+2+2+2 = 10 ;D )

Furthermore this can lead to ridiculous situation, where a failed Influence throw to aquire a stub revolver fails, but the influence throw after this one in the same shop from the same acolyte can requisition a melta (success in second throw).

I think there's a simple way to handle this; do away with tests for acquisitions that are pretty trivial. One way I've thought of working it out without too much difficulty is this. Subtract 20 from your influence. Now, any acquisition you wish to make that's below Inf-20, you can take without any test.

For example, say your influence is 50. Take 20, you've got 30. That means anything -30, -20, -10, etc, you can just requisition outright. It still costs an acquisition, but there's no need to test; you're assumed to be wealthy and affluent enough to be guaranteed success.

That doesnt solve requisitions DURING the game.

There should be a fixed amount of Money (even if it is abstract) during a mission.

If its not the case, then in one encounter they can "buy" something with Rarity X, in the next they can't and the third time they can again due to a good roll. That doesnt make much sense. If I allow them a threshold to buy freely from...my acolytes will buy ANYTHING - you read me right ANYTHING. Maybe Rogue Traders "Ambition knows no bounds" coined them too much.

It would be better to have an abstract pool to draw from.

You wanna buy that Bolter in the "Gunmetalls Finest Outlet Store" - no problem, but it will reduce your Abstract Money Level from 6 to 5. No ? You rather take the stub revolver ? That doesnt lower your Money...at least not buying just one.

And this Money Value you have during the mission should be requisitioned BEFORE the mission (a bag full of thrones)

I'd say it'd be easier, if a little less versatile, to just allow multiple acquisitions beyond the limit of three, but reduce overall influence each time, to represent dwindling wealth. I'm just not convinced the system needs to get any more complex.

I just want to avoid those situation, where good A is not requisitionable (Test failed), but good B later on (Test succeeded, even if more difficult) is, even if you carry the same money with you..

Acquisitions should have both hard and soft limits. A soft limit is the base number of Acquisitions you can make during any period of downtime, or at an opportunity that arises during an Investigation. A hard limit is to simply yet firmly indicate that once an Acquisition Test is failed, no more can be made during that period or opportunity. Failure to acquire a stub gun followed by a success to acquire a power sword is a prime example of "fishing."

There should be a clearly defined order of process.

1. The Player makes a list of the items he/she wishes to acquire. This list may be prioritized any way the Players chooses, but it's best recommended that easier Acquisitions be first, followed by a more difficult Acquisition, and so on. It is during this step that the Player determines the availability modifiers of each item.
2. The Player uses any Skills or Talents he/she feels would aid in making Acquisitions more favorable, items easier to obtain or cheaper. Successful Tests at this juncture result in either a bonus or penalty to each individual Acquisition Test (such as Commerce in Rogue Trader, where the PC might earn a cumulative +5 per DoS). Once these modifiers are applied, the Player may rearrange the list's priority as he/she chooses (still recommended easiest to hardest). Failed Tests at this juncture do not preclude an Acquisition, they simply make the items more or less available (expensive, low stock, etc). This "haggling" is done once for each item (unless the Player favors his/her chances without a bartering process or doesn't care about being subtle).
3. The Player then makes an Acquisition Test for each item, one-at-a-time. If at any point the Player fails an Acquisition Test then no other Tests can be made. This is a clear indication that Influence is currently waning, funds are low, "dummy" accounts are not accessible, items are simply not available, etc.
4. Successful or failed Acquisition Tests could leave an impression (for good or ill, whether successful or not) on the merchant/personage from whom items were obtained. I've been using the Reputation mechanics from Rogue Trader. A PC might choose to use Deceive as a "haggling" leverage; particularly successful Deceive Tests might make the Acquisition more favorable, but something about the transaction may not "sit well" with the merchant (I might make the haggling Test Opposed), so I tell the Player to make a note "Reputation: Con Artist, Footfall, -1." If the PC chooses to use Deceive in that venue at a later time, he already has a suspicious reputation, so particularly successful Tests may earn him a healthy bonus to the Acquisition Test but an increased negative reputation, say -2, while a particularly poor Test might result in the merchant kicking the PC out of his shop and refusing to do any business with him in the future. Reputations range positively from +1 to +3, and negatively from -1 to -3. Good reputations (positive) may result in lingering good will- maybe a +5 bonus to a future Acquisition- while bad reputations (negative) may result in cover being blown as the local enforcers are contacted regarding the PC's background and intentions, or the merchant simply passes the word down the line about a "shifty and tight-lipped" off-worlder...

The GM can use Subtlety to make Acquisitions even more or less difficult, depending on whether the PCs are still inconspicuous or are suspected of being "involved" with the strange goings-on...or if they happen to wander into a blade emporium funded by the very cult the PCs are Investigating...

All of this shuts the "fishing" tap off at the source.

EDIT_

I should note that this is the framework I've been using. To "curb" overlong shopping excursions, I impose occasional Influence reductions for particularly poor Acquisition Test results (-1 for 3 DoF, -2 for 4+ DoF), though I do allow the character to then have some item, not necessarily the exact one he/she wanted, but something similar/lesser to show he/she received something for their time and trouble, and that they either stretched the limits of the resources or were swindled/taken advantage of in some fashion.

Also, if a particular venue is "stressed" for supply, I impose cumulative penalties to each Acquisition Test after the first: -5/-10, depending on region, technology level, etc.

This sounds like a really good idea

I think I will implement this for the future !