Well the problem with the table/read is, that it you cannot control time and pitch individually, as far as i know..... So when you touch the pitch of the phasor osc, you are editing both the pitch and time, which is not a good thing in most cases, but in some cases it works well(like for eventidish harmonizer delay like effects, some sampling situations)... But for timestretching, reverse audio etc. It doesnt work well, because of the pitch issue that you have faced. You basicly have to pitch the sound down like 8 octaves or something for the audio to be in big enough chunks to actually here that it is played in reverse... But then you wont hear the original pitch of the audiofile, it will be totally unrecognisable.... We need a table/read with individual inputs for time, pitch and position for that. That would be really awesome and woud give a lot of new opportunities. I think Alcehmy from Camel Audios functions for stretch, reverse, etc. are very desireable. Alchemy is the best and most controlled VSTI I have ever tried and have beeen trying to build something similar in Axoloti, but get stuck when I get to the time & pitch issue... Maybe it is possible(like reverse audio), but havent found out yet... But for Aclhemy pitch is editable by itself and the time is editable in percentage and position also has its own knob.... That would be sooooo good to have in Axoloti.
But I guess it is up to the developer to decide which direction to go. Or maybe they have an idea how to accomplish what you are trying to do, from what we have now. We could ask @johannes and maybe also @thetechnobear if this is something they put some thoughts into?