I wasn't going to respond to this since I'm against the use of flags but didn't want to defy the OP by arguing about it anyway. However, since the discussion is underway regardless...
My own bike is low (I can balance on both hands at traffic lights) but I have also done a little distance on trikes borrowed from Laidback & friends - about 600 miles off the top of my head, including central Edinburgh rush hour commuting. Never with a flag.
Wouldn't want you to get cross.
But... I have to report that since I hoisted the flag the amount of abuse and near passing has definitely reduced. I think it weeds out the desire to teach me a lesson for 'being invisible', and the height sense secures a tad more roadspace.
My experience is that small changes in the way you ride can make overwhelming changes to the way people drive., so I'd be very interested to know whether because you know you're flying a flag you ride differently on the road in even the smallest way.
This is the only way I can readily account for it changing the behaviour of motorists, since (to lay down some anecdata) I would say that it's often impossible for drivers who pass me to do so leaving any extra room. However if I rode more passively, I've no doubt there would be a lot more close passing. Just a suggestion.
They also seem like a good idea for very low recumbents when riding in city traffic, for edge cases like the overtake-into-space-occupied-by-the-bike manoeuvre....
I've never felt the need for one on the Streetmachine, but it's as high as the average car, so I wouldn't expect to.
If you think about it for a minute, the extra inches of the Streetmachine will often be overwhelmed by extra inches (or feet) of following vehicles. As I often remark to concerned parties in central Edinburgh, they are no more visible through the 4x4 or van behind them than I would be.
This type of vehicle makes up a sufficiently high proportion of city traffic that I think we would see these same "edge cases" impacting normal riders if it was a genuine threat. Yet I've never heard of someone being hit in town because somebody overtook into the space they were occupying. (That's not to say it doesn't happen, just that it's not worth worrying about IMO compared with the "usual suspects").
I think flags on some recumbents are useful - particularly on Trices - when you are in heavy traffic. It allows cars that aren't directly behind you, or cars in front of you, know that you aren't an empty space.
All of the examples I can think of are instances where the flag might mitigate bad cycling. For example, if you are riding in heavy traffic, you may certainly end up creating a "gap" beside a junction when everything stops. In this case we can see why it's important to ride in the *right* tyre track such that oncoming right-turning traffic can see you. If someone commits to turning into the percieved gap at almost 30mph, they probably won't see the flag in enough time to stop anyway.
That's not to say that having the flag does you any harm,
providing you ride as if you don't have it (key point alert). It's just important to understand that the edge case where it saves you is extremely edge indeed (like having a third tail light which saved you compared with having only two, IMO).
On the trike, I am normally going slower than the rest of the traffic. If a car moves out to pass me, it is quite possible that a car following him is completely unaware of me until the first car has moved out. Due to the speed difference that can create a hazard that will not exist with a low car.
Imagine that instead of a trike, you are actually a set of cones around a work pit. Urban cones are often small - certainly no higher than you are - and will not be seen by any following vehicle. When the first car moves out to avoid the pit, will every subsequent vehicle ignore this, drive straight on and crash down to ignominious destruction?
No - in fact I'd argue if anything that being an "unknown hazard" forcing the lead car to move far out is actually *better* for your safety than if following vehicles know they are being held up by a cyclist.
Similarly - imagine you are being overtaken by an HGV. I can easily accept that in some scenarios the driver might be able to see the flag and not you, but think about the two alternative courses of action: the driver can't see if he's past you or not, or the driver can see a bit of flag and has to estimate whether that distance is further back than the end of the trailer. In which scenario might they err on the side of caution?
Another difference - there are places I can filter on the trike. This means I put myself alongside other vehicles where a low car will not, and hence possibly in an unexpected way.
Yes - going back to the previous example, you (by which I mean, "someone") might choose to ride up the inside of a left-indicating HGV. If the driver doesn't have the appropriate mirrors, it's easily possible that they might see the flag but not the rider. Yet this scenario is predicated on the cyclist choosing virtual suicide.
What concerns me about this idea is not so much that a flag will make filtering more dangerous, but that having a flag causes people to filter much more dangerously. IMO, there should be no scenario while filtering that relies in any way on the traffic you are overtaking having seen you - that's horrendously risky. (Even though in reality, so many people ride like this every day that we must accept it to be only *relatively* horrendously risky...)
So, while I flinch from the idea of telling you I don't think anyone should have a flag, I think on balance (as with universal use of hi-viz, helmets etc) it comes with more baggage than advantage for you and for everyone else.
(Why does everyone ask me about flags? It's not because they haven't seen me but because they have seen recumbent riders who've gone that way).
