The problem is that the “lack of tools” is just too big to overcome. If all your sound comes off a pair of speakers, there is no realistic chance to reproduce the depth and complexity of a real room, where you can perceive immediately which sounds to prioritise. You’d need a full surround system, then you’d need a webconf instrument that can interact with it properly, then you’d need the equivalent for vision - a single screen is just too small and depth perception is nil. And you can’t really get there incrementally - it will continue to suck hard until you have every element sorted out. And of course there will always be “i can’t hear you now, can you hear me? Oh, she’s disconnec— ‘I’m here everybody, sorry Jerry, what were you saying?’ - I was saying that — ‘I can’t hear you Jerry!’ “ etc etc