One figure or a whole crowd for a crash diorama?

I did somthing similar a couple of years back with a 1:35 flatbed lorry jammed under a low railway arch, and I went the opposite way on the figures question. I kept it to just the driver and his mate standing on the footplate staring up at the arch, and the emptiness is what makes it work. Crowds can easily turn a quiet, embarrassing moment into a street party, and it stops reading as one poor bloke's very bad day. That said, your mate's pointing-at-the-sign idea is genuinely funny, so maybe split the difference. One figure, positioned well back from the truck near the warning sign, mid-stride and walking away like he wants no part of it. Distance keeps the focus on the driver while still telling the viewer the world noticed.

For figures, the old Preiser sets are decent value, and a bit of gentle repositioning with a hairdryer and a fresh coat of paint sorts most stiff poses. I painted mine in muted tones, greys and browns, so nobody upstages the crushed roof.

One thing I would add is the aftermath detail rather than more people. A scatter of debris, a dropped clipboard, a cone knocked over. Those little touches did more for my lorry scene than any extra figures ever did.

I built a 1:24 delivery van wedged under a car park barrier a couple of years ago, and what sold the scene for me was not the people at all, it was the aftermath details. Skid marks on the tarmac made from thin Tamiya smoke painted on, a discarded coffee cup, the van's roof panel sitting on the ground with the barrier light bent over it. My figures came last and honestly they were almost an afterthought by the end. So if you want more in the scene, I would look at props before bodies. A broom leaning against a wall, a wheelie bin moved out of the way, some traffic cones the driver clearly ignored. Those read as story without turning it into a crowd scene, and they work at any viewing distance, which figures often do not.

On the pointing-at-the-sign idea, I would be careful with the pose. Gestures like pointing are hard to pull off in miniature because the hand is a couple of millimetres of plastic and from most angles people cannot tell what the arm is doing. A figure standing with hands on hips looking up reads better and takes less conversion work.

If you do add anyone, a single workmate type leaning on the crash barrier with a phone out would be my pick. One extra body, and it suggests the whole 'wait till the gaffer sees this' thread without needing a queue of onlookers.

One pedantic thing to check before you call it done: the number on the height sign. If the truck is, say, 3.9 metres in 1:24 terms and your sign reads a clearance the vehicle could actually have squeezed under, anyone who knows trucks will spot it imediately. Measure the arch on your diorama, knock a sensible margin off for the sign's own tolerance, and write that figure on it. I did a wrong-turn coach scene a while back and spent an embarrassing amount of time with a calculator and a ruler getting that sign plausible, and it is the detail people zoom in on first.

I'm pretty new to this and the one thing I got stuck on with my little crash scene was making the dented roof look right. I tried squashing a bit of thin plastic sheet with my thumb and it just cracked. Ended up pressing it over the back of a spoon that had been in hot water for a bit and that worked way better.

Does the same trick work on the moulded truck roof, or is that too thick to shape once it's glued on?