Forged in the dark : one large clock, or stacking small clocks ?

If the bad guy hears rumours about someone asking the question, does anything change?

When a clock fills in these contexts that should indicate that something needs to happen, and that something likely requires PC response. So if it isn't going to significantly impact the PCs until the third clock, it may as well be one big clock with stuff happening in the background as it fills. But if each clock has an impact and the PCs can do something to impact future clocks, stacking makes sense.

Regarding handling the consequence: it definitely depends. I'll sometimes use a clock if they're trying to overcome some major obstacle, so filling it means that they have less to deal with and that's probably going to be an RP exercise. But most of the time it's going to result in a change in position, or a need to resist something, or even a material change in their crew's territory that requires some response. In the example above, especially for such a large clock, I'd probably have the consequence be something like the bad guys invading their territory targeting the PC asking questions, which requires more than a mere change in position to resolve. That could involve a full-on heist to thwart.

I don't think there's one answer to that. To me it depends on the context of the clock and what's your plan for pacing. Also it will be part of your style that you just have to find for yourself, what works for you

(Cyberpunk examples)

• Tripping guards suspicion on-site: one small clock
  Consequence is not an alarm yet but from now on everything that has to do with guards can have lower position
• Tripping alarms for the whole building: bigger clock
  Or don't set up such clock at all if everything going completely south doesn't fit your overarching plot plans
• Mafia responds to characters asking around: small clock
  They have reputation to uphold, they can't have someone nosing around in visible way Consequence:
  • someone who said something gets in trouble, making others harder to work with (lower position)
  • Mafia learns who they are (if that would be serious problem for the whole run, I'd make it a bigger clock)
  • They get set up and have an unplanned meeting with a bunch of enforcers
  • It gets so obvious that they get contacted by this group's opponents and the situation is stacked that characters either comply with demands or will have very hard time completing the run
• Police/corp responds to characters doing runs against the corp
  • If you plan the corp to be present in the plot, make it a big clock to fill it after a few runs
    • Or make it small to force the characters to manage their footprint from the early stage (lower position when doing things the corp can piece together)
  • If it makes sense that corp would first send police after them, make it two small clocks
  • If you don't care about the corp, make it a short clock, to hopefully resolve it during this session
    If they manage to not fill it, after all, keep the clock for the future. The next time you feel it's going too well for them, you can fill this clock instead of more current one. Suddenly bringing old grudges into the mix

So depending on what you want to do it's either bigger or smaller clock, with consequences either in fiction or mechanical

It's been a while since I read the rulebook, but what I remember was that completing a clock causes a thing to happen. So I imagine smaller clocks are generally better for incrementally upping the ante, e.g. first the guards know someone's here (making your position default to risky), then more arrive (meaning stealth is out and you need to run or fight).

Can we stop with the clockshaming? All clocks are good clocks.

I don't usually use any kind of pre determined timing that would be measured with a clock. I use story beats to introduce new ideas and events in the story so it always flows naturally based not just on what I have my NPCs doing in the background, but also what the PCs are doing in the foreground. I might have things that are timed out in weeks or months that can be triggered by the party resting or training or whatever, but this would only be because they've triggered an event like an incoming invasion force or something as it falls along the way the story has progressed.

Let's say the players are asking questions they shouldn't be asking about...

If the players are asking questions that are metagaming, I direct them back to the game and what information their characters have and what questions they would have based on that. If the characters shouldn't be asking the question because of some in-world reason, such as the characters they are talking to are loyal to a king who finds such talk treasonous, well then it would just play out in-game taking those things into account and the party may be pushed into a fight or taken to jail.

I employ as much improv as I write things down ahead of time. There is no way you can account for every single thing that every other player at the table may do, so I think it's best to leave a huge degree of vagueness until things start happening, so you can just make shit up on the fly for things you couldn't prepare for.

Time is relative. Especially in story-telling.