Thanks for the elaborate feedback! It's super helpful and actually the main reason we are making the game available at such an early stage.
That feeling of 'hm, something doesn't add up quite right' yet about the flow of things is something we want to balance out with feedback from the community. As you mentioned some issues with the input system - maybe that could be the 'missing part' that kinda puts you 'out of the flow'? You're also right that there's not a lot in the way of tutorials or introductions - this is something we want to tackle soon. We have not done so, however, because we haven't settled on a control scheme yet. This is a topic we've put a lot of consideration into - one of the things we've decided was that unarmed and armed controls are going to be different. This also ties in with your observation that there only seem to be ~four moves: This is actually due to the fact that there are only really three types of different boxing attacks - straights, upper cuts and hooks. They may target different body targets and be executed with either hand, and their angle may also change slightly. But they are generally limited to pretty narrow angle ranges that also do not really transition very fluidly. This is of course because we for now take grappling, throws, choking, clinching and elbows out of the equation.
Contrast this with weapons combat, where swings can transition fluidly in the upper 240° - which you can actually do with our armed combat. That is one part of the question why we do not adopt the same attack scheme like M&B. Another is - we do not like essentially just 'watching' attacks unfold after the 'charge' was 'released' in M&B. It creates a disconnect between what is happening on the screen and what you actually do/control as the player. Another issue is that it causes a delay to attacks. We instead work with a 'guard' system - think about the 'charged' state in M&B, e. g. a raised arm, ready to strike. Actors would mostly be in a 'guard' state that allows for very quick strikes from that position, but also telegraphs the angle. Strikes from vastly different angles take longer in contrast. It's all very early, but there's been great progress. You can see a gif of it in action here: https://twitter.com/KinstrifeGame/status/1034078220918943744
By the way, have you explored the different attack control schemes (e. g. drag & release etc.) yet? Any thoughts on that?
Regarding footwork - Do you mean general combat movement or how footwork aligns with attacks? The former is actually because we currently do not have proper animations for combat locomotion. You'd be surprised, but the majority of combat locomotion uses crossed feet... Which can be especially fatal with physics based combat and tripping ;) It's very high on our list of graphics-related things we want to get replaced, though!
As for speed - that's a good question and may even be up to personal taste in games. As you said, reactions in real-life are very instinctual - trying to put up one's hands to protect the face is a super important instinct. But also reactions like 'uh oh, I should probably duck now and avoid that hook' are very instict-based - which is not an issue in real-life because that very easily and naturally translates to movement of the body. But with games, it has to go the route of remapping that to a game control scheme and clicking buttons. There's certainly going to be a noticeable delay and it's something we need to keep in mind.
Again, thanks for the elaborate feedback and we're super happy to have discussions like that! But for the sake of posterity & overview, we'd love to bundle discussions like these over on our forums. So it'd be great if we could carry on there :)
EDIT: Woops, just saw you've already signed up and posted there :)