Already got 500K with Switch Pseudo-TASing. I caught enemies on most fields, grinded on Fields 17, 37, 38, 41, and 48, and 1/2 of 5 (exited with 2 minutes remaining). I entered 57 with 85,000 points. I implemented a new strategy of skipping Field 56. My Field 57s are VERY good with scores better than 22,000 per life. They would be very nice to get in a real run.
500K is 100K better than what the range for the current record is. I thought the potential score couldn't be much better than a low 400K, but it's actually much higher. Very encouraging. Plenty more margin for error than previously imagined.
Are High Score runs too long and tedious for you? How about Deathless High Score Runs? Still too long? Here's an idea: High Score IL Runs.
Here's how it works: You do a High Score Run on each of the game's 49 Fields. Your best score on each Field adds up to your Total. The different Exits don't matter, nor do extra lives, since they don't add points to your Score. Your PB on each Field adds up to your Total. Also, you must go through one of the Field's Exits (doesn't matter which) to make the Score count. As long as you don't lose a life or run out of time and go through a door, even with 0:00 on the Timer, it counts. I thought about having separate LBs for each Field, but there'd be a lot of unbreakable Ties, so I just decided to have one added-up Total for each player, as well as all WRs.
You can't just grind- only your best score will count obviously. So if you get 5000 and then 5200 on the same Field, you will only add 200 not another 5200.
This challenge will make the tedious High Score challenge much more palatable- and it will open it up to the rest of the game outside the Longest Path.
In order to count IL Scores on something that isn't the Vita or the PC, if your starting score isn't 0, subtract the starting score from the ending score. All versions are viable for this challenge. Since the Switch has savestates and rewinds, all ILs must be fully recorded from the pre-Field number blinking to the Exit.
Umihara Kawase recently came out on Japanese Switch VC in the Super Famicom section. It's the exact same game as the Super Famicom version, TNN logo and all. The only difference is its portable and has savestates, rewind, and the ability to change the controls, all of which come standard on Switch. I still feel the Vita version is better- it has a Field Select, it natively saves your high score and all your times (you need savestates to do that on Switch), and it's on a physical cartridge, where the Switch version is impermanent and digital. You can expect it to no longer exist shortly after the next Nintendo console comes out. In addition to that, I also feel that the Vita generally has the better Umihara Kawase games than the Switch does, but you probably already know that.
You can get it by starting a Japanese account (easy) and subscribing with your own local account, which covers all accounts for that particular Switch. If you only have a Japanese account, you can get yen by paying for online codes on import sellers like Play-Asia.
I'd still recommend the Vita version, but far more people have a Switch than a Vita or Vita TV, and the Switch version is still a pretty solid and cost-effective option.
The game came out on Switch recently. The Switch version is on the online VC, so it's not the improved 2015 version. It does have savestates and rewind though- making it good for testing the limits of high score runs. Already, I've done one- a rough, could've-been-better test run on Switch with these pseudo-TAS features bursts open the possibilities. The new high score is within the 450K range and it could be even better. I'm not expecting anyone to reach that in a real run, but it does make a low 400K much more beatable.
After the run, I savestated, meaning a great feature from the Vita version, the ability to save your score, is maintained. Next time I try for a better pseudo-TAS, I'll have the score to beat right on the screen, giving it something more of a personal significance. It sure beats reverting back to 6158 each time like on Super Famicom.
The Final Version of Star Fox 2 is now available- on SNES Classic, Switch, and even on an original cartridge. It's better than the Beta version, and also more challenging- your charge shots don't home in on enemies anymore. Cleaning up on all those missiles won't come so easy- it's not a matter of merely having the enemies onscreen anymore, and your bombs are limited. There is, however, a way to make them not limited.
This is best done when there aren't many bases left on planets, and no Battleships left. And on Expert, before the next three Bases appear. Enter a Planet, collect a Special Item or two, then exit. Hardly any total time will have elapsed, and you can do this a few times to have the maximum of 9 of whichever the item you last picked up was. If you have 9 bombs, you will quickly and easily beat Star Wolf members with just a few bombs each, which reduces risk and gives you ample time to stop attacks on Corneria. Bombs are also useful for Bases that have four separate targets in one core, and on groups of Missiles and Flotillas. At any time, you can also collect a single Heart or Shield to get one extra and change all of your Items into that thing. Even Expert Mode, while not effortless, will become much easier due to how much less scarce Items become. Missiles are easier to stop with Bombs, full-heals are plentiful with Hearts, and tight flying sections will become safer with more Shields.
Any game with a learning curve as sharp as Umihara Kawase will throw off 99% of players before the first Game Over screen. There are any number of little tweaks possible to address this, but I'm not sure I'd want to implement any of them. It's ok for a game to be niche.
But to answer your question: speaking for myself, the number one gameplay factor that nearly threw me off the games altogether was the random enemy spawn. It leads to endless bullshit, unavoidable deaths during the first few dozen hours, until you develop the sixth sense required to predict when and where the random thing is going to happen. If they wanted to randomize enemy layouts, have six possible spawn points per level and have the game randomly pick four spawns when the level starts, or some variation on this. No real-time spawning, though - it just messes with the flow of gameplay, and provides little benefit. I guess it was included to dissuade loitering but the level timer would've been fine for that.