Any pointers to the ideas that have been discredited? This is a technology I dearly want to work, and I had bought into the narrative that it doesn't "because we haven't invested enough" (compared to offshore oil and gas production).
You can broadly group all tidal devices into several simple groupings and for each of those there were a number of initial prototypes and all of those did not make it past the POC or initial deployment. Anything the future is likely to come up with will be variations on those themes. That won't stop people from trying and it won't stop future subsidy grabs because this all falls under the 'wouldn't it be nice if it worked' heading. Yes, it would be very nice. But no, it doesn't work and if you're not going to come up with something radically new you are better off reading the literature on what exactly made it not work the first time.
Those who don't know history...
Here is a very nice example:
https://teamwork.nl/category/teamwork-technology/
I'm sure they pocketed a ton of subsidy by now.
In the same vane (pun intended): rooftop windmills, vertical axis turbines, kites, sound to energy conversion and on and on.
This is really key, there just isn't that much energy available in a small enough area for utility scale production. I would add the caveat that for small scale production, some schemes could actually work quite well, like some kite schemes. But no one is investing in small scale, all the development is for multi megawatt systems.
Really, marine engineering is an art, and anything that isn't solidly anchored is going to be the oceans plaything sooner or later.
By analogy, just because early attempts to build a concept does not work, can you prove future attempts won't work? As far as I know, the energy potential in the ocean is not disputed, just the initial attempts to build the machines have not been successful. Many attempts at flying machines were unsuccessful before the wright brothers.
Tidal energy has been tried time and again and nobody has managed to make anything that would work at scale.
Your method of analysis is assuming we exist in the best of all possible worlds. "Flying machines have been tried time and again" is valid right until wright brothers. "Electric cars have been tried time and again" right before Tesla enters the market.
It's one thing to say "roof top wind" and you can look at the loads an average roof can support, realize it will be a small swept area at a low elevation and conclude that roof top wind couldn't power the building its mounted on. There is no way to write off tidal with the same sort of first principles analysis, the power is there, its a matter of design whether or not it can be captured efficiently.
Highly annoying to see ideas that were - solidly - discredited decades ago come back in a slightly different guise for a re-run.
Based on past performance I predict that this setup will right up until the first major breakage, after which it will be uneconomical to even salvage it.
The temptation of the tides is easy to understand: the power is immense, it is essentially the power inherent in the gravitational attraction of the moon orbiting the earth conveniently transferred onto a moving fluid. So on paper it is trivial. But that's where it stops being simple. The forces being immense is also a force that tries to destroy your gear. Corrosive salt water is the nastiest working fluid you could pick outside of acids. The wind & the waves combined can really stretch the limits of engineering (I've seen a minor bit of water hammer effortlessly flip an 8 cubig meter concrete cube right across a dam, like a 2x2 lego brick only slightly larger and heavier).
The 'head' compared to most hydro installations is limited (though in some locations it can be of usable height), the flows are hard to trap without interfering with the rest of the environment and so on.
As a 'mere matter of engineering' it is probably one of the hardest jobs you could tackle including going to space, the ocean is an extremely dynamic environment and near the ocean shore that is even more true.
And finally, to be economical such an installation has to work for a very long time after initial deployment because it is very hard to work on once deployed.