Look. In theory, RTWP is superior. I have no problem admitting that it is more realistic in the sense that everyone is moving and doing at the same time. There is no, "He moved, then bad guy 1 moved..." as if an enemy is just waiting for you to move 30 feet and attack him. It's also faster paced, if all the characters do what you want them to.
In fact, I'll wager that there is not a DM in the world who would say they WOULDN'T prefer to have a way to run battles in TT using RTWP if it was possible. The issue is the execution of it. Having all enemies and allies moving at once is chaotic and difficult to track.
I've even thought up ways a video game could do something quasi-TB/RTWP where you enter actions for all your party, telling them who to move towards and attack and with what, then the game unpauses, all characters move at once, do exactly what you said, and then it autopauses because the round is over and a new round is beginning. So the main difference is it is still round based, but instead of initiative order everyone moves and does their actions at the same time. Initiative still comes into play, but only if someone's attack would impact someone else's.
The problems for me with RTWP is scripts and I can never tell when a character has completed a move or attack that I told them to do or whether my weapons or spells are effective or not. So I would tell Minsc to shoot at someone and I would look away for a sec and wonder if he had shot yet or not. Just as I'd tell him to move somewhere, I'd see him pull his bow back like he was about to shoot, and he'd cancel the move and go where I told him. If I'd known he was going to shoot that second, I'd have waited a bit longer.
Or, in the case of the final battle in Throne of Bhaal, I had to pause a thousand times and look at the battle logs. Were my characters simply not hitting or were they doing no damage because of damage resistance or some damage immunity or something. Or I'd have to pause because I'd notice Imoen wasn't doing anything AGAIN! She'd cast a spell and just stop. I tried a different script and she'd be cued up to cast a spell only to stop and just start firing her bow which would do no damage because whatever it was she was shooting at was immune to her bow's damage, or whatever. It was too hard to figure out what the frick was going wrong.
I don't get that in Turn Based. If a spell doesn't work or a weapon doesn't work, I know after it happens once. My characters don't spend round after round swinging pointlessly before I, the player, realize something's not working right.
It was always things like that which frustrated me, that and party members wasting spells. Shoots a goblin with lightning bolt when it has 4 HP. Stupid. Waste's Magic Missile on a wolf with 2 HP. Why? My fighter was about to kill it.
So if they did a quasi-TB/RTWP, somewhat like the XWing miniatures game, that might be a bit more acceptable, but it would still be hard to determine what went wrong during a particular round and it would still take about as much time as just doing Turn Based mode. But it might be more fun and realistic.
First Round. Player selects Shadowheart and commands her to attack Goblin 1 who is within 30 feet of her. So, when the game unpauses, she will run towards Goblin 1 and attack with her mace. Player then selects MC and chooses to have the MC move to behind a rock, use Cunning Action to Hide as a Bonus Action, then come back out from behind the rock and shoot Goblin 2. Player then chooses Gale and orders him to cast Magic Missile at Goblin 3. All 3 missiles will launch at Goblin 3. Finally, Player selects Lae'zel and commands her to run up to Goblin 1, because Goblin 1 is the only one within 30 feet. She will use her Flaming Sword Everburn to attack him. Commands all locked in, Player selects Begin Round.
All characters move at once on the battlefield. While the player was locking in commands for the characters, the AI was locking in commands for the Goblins. So, the computer calculates who is moving where and doing what. Goblin 1 realizes it is in trouble and retreats 30 feet and then fires a bow at Lae'zel. As he moves 30 feet away, Shadowheart and Lae'zel both move 30 feet towards him. Because the computer determines that they will not reach Goblin 1 with melee this round, it auto-switches the characters to ranged weapons and has them fire at Goblin 1. Shadowheart fires first, then Goblin 1, then Lae'zel because that was how they fell in Initiative Order.
Meanwhile, MC was shooting at Goblin 2 who was shooting at Shadowheart and running 30 feet to get to high ground. Goblin 2 came in Initiative Order after Shadowheart and MC, so Shadowheart fires at Goblin 1 before Goblin 2 fires at her, but MC fires at Goblin 2 before it shoots at Shadowheart also. So, let's say Goblin 2 is killed by MC before it can fire in Initiative Order. In that case, it never fires the shot.
Goblin 3 and Goblin 4 were both going to shoot at Gale, but they were charging at him. So they run 30 feet towards the PCs and shoot at Gale. However, Gale won initiative, so his magic missiles will fly at Goblin 3 before it gets a chance to shoot. Gale kills Goblin 3, but Goblin 4 fires and hits Gale. All these things happened simultaneously. Game pauses.
So it is kind of RTWP except that it is round based instead of each move needing a specific Time Cost. A dagger wouldn't swing once per second versus a greatsword swinging once every 4 seconds or whatever. It would maintain the Turn Based action and bonus action count.
Now that might be fun and might work well. Then you could see your choices played out in realtime as everything moves at once and attacks, and it might not require Larian to set the game up with the whole attacks per second thing. Still, there are other issues with this that I won't go into right now. This is already long enough.
It's interesting you say this because this is what I myself have often proposed as a good approach. And in fact I did play TT D&D with a DM once a long time ago who experimented with something like this. The system itself worked great. The problem was always the players who, coming from their D&D background, constantly wanted to change their actions once they saw what someone else had done. That became annoying and frustrating, both for the DM and for those players like me who were vested in the simultanous-play system.