Given only two speakers, the best you can try to do is aim for a best-effort compromise across multiple listening positions. Someone, or everyone, will experience time- or volume-related inaccuracies.
With more speakers, and signal processing, you have potentially more options and can therefore achieve a better compromise (e.g. Trinnov), but no one has yet produced a system that uses active cancellation to produce multiple sweet spots that do not interfere with each other. Alternatively, you could go with a single "point-source" speaker.
As a thought-experiment, you could have two speakers so far away from the relatively small physical area where all listeners are located, so as to have minimal SPL loss or differences in timing. But that's probably not possible in your home unless your living room is a TARDIS.
With more speakers, and signal processing, you have potentially more options and can therefore achieve a better compromise (e.g. Trinnov), but no one has yet produced a system that uses active cancellation to produce multiple sweet spots that do not interfere with each other. Alternatively, you could go with a single "point-source" speaker.
As a thought-experiment, you could have two speakers so far away from the relatively small physical area where all listeners are located, so as to have minimal SPL loss or differences in timing. But that's probably not possible in your home unless your living room is a TARDIS.