So many of the posters to this thread want to improve safety ... but only their own safety, at the expense of the safety of other pilots.
For example, englishal (as the most recent only of these) wants to restrict my glider flying to 3,000 ft, which puts me at the risk of around 4 field landing per hour. Is that an acceptable trade-off? Or I should fit a transponder, which restricts my time in the air to 3 hours or so and will cost 30%-50% of the hull value of my aircraft. I note that he doesn't intend to change the way he flies to improve the safety of others.
There is no absolute safety in flying - it is a matter of minimising risk to the lowest acceptable extent. I can think of many ways to improve safety which would be unacceptable because they would stop others flying. As an example, how about excluding from class G airspace all aircraft which have restricted visibility from the cockpit? Gliders would keep on flying, but many powered aircraft would not. Clearly it is safer to be able to look out as far as possible, but in my view it is not acceptable to propose restrictions which would effectively ground or radically restrict the operation of aircraft which currently fly perfectly legally.
Within that constraint, what could be done?
Mode S, ADS-B etc are out because of the practical (power and space) and financial constraints.
Something FLARM-like could work if it were permitted without certification. Allowing the fitting of uncertified transponders (as in US gliders) would help, but wouldn't be as good as a universal, low power consumption solution because many users would need to turn the transponder off outside high traffic flow areas to conserve power. Certification seems to multiply costs x 10, which is why an uncertified solution seems to me the only possibility. I suspect that if enough gliders fit FLARM, it will be worthwhile other pilots adopting it as removable equipment (like a hand-held GPS). I'm not pushing FLARM, merely using it as an example of the kind of cheap solution which might be workable.
The cheapest solution is for all of us to work on lookout. If I haven't scanned the entire field of view (that's about 120 degress fore and aft, and 240 degrees side to side) at least once every 30 seconds I slap myself on the wrist. Some pilots would say that's barely adequate. Do we all set appropriate standards for our lookout? Do we all achieve them? This won't avoid all mid-air collisions, but then neither will any of the other solutions proposed here.
Finally, it's worth noting that the genesis of this thread is the collision of two powered aircraft on approach, both talking to ATC. No-one has proposed ways of reducing that (clearly real) risk, but the accident seems to have focussed minds on lower risks which might be avoidable if others changed the way they flew.
If it's relevant, I don't think I've ever flown so as to require a non-glider to manoeuvre to avoid me. I have taken avoiding action on a number of occasions when a powered aircraft has flown straight at me, on the assumption that the pilot hasn't seen me. This doesn't surprise me, given the restricted field of view, my lack of visibility and the number of things in the other pilot's cockpit which require attention, so (rules of the air notwithstanding) I make it my responsibility to look out for such situations and avoid the danger. I don't find this risk unacceptable.