I do understand your point about the appeal of game options that make the game quicker, less predictable, etc., and I agree that tournament/pro rules are not the only measuring stick. So I have taken the ones that would feature frequently with the exception of the tie-breaking by initial seat assignment (part of point 6), namely, 1, 2, 4, and your other suggestion (single-hand game).

Points 3 and 7 are exceedingly infrequent, irrespective of rules, and it's difficult to justify putting in the time. Points 5 and 6, or some aspects of what you suggest, don't fit into my understanding of BGA.

Details:

I implemented 1. Straight head bump (only ever single-Ron) or double-Ron possible but only first in turn order gets bets/bonus (Tenhou default I think?).

I implemented 2 (I agree that this is much more common than even single Yakuman hands if players don't control Dora tiles).

I am not implementing 3. To me it is irrelevant if players are familiar with the idea if it (practically) never happens. Also, this would be much more important in longer games or tournaments rather than in the typical BGA scenario. Same with the sky-rocketing variant.

I implemented 4 (both options).

Re 5. Oka/Uma are not meaningful here since there is no continuation or cash payout in BGA. So calculating a linear transform of the original scores is just adding more confusion for beginners (and the rules aren't exactly beginner-friendly) while not changing player order (which is the only thing that matters for ELO afaik). The reason they are in the EMA rules is because EMA organizes tournaments. For splitting ties: I personally don't like largely arbitrary tie-breakers and much prefer to keep the ties. There is also no meaningful way to do this for the single-hand game, leading to further options/rules divergence, which I want to avoid.

Re 6. I am not sure I understand your point. There is nothing obviously obscure or unrealistic about continuing to play until there is a single leader (which is the definition of sudden death). While I am familiar with the rule, I never implemented minimum target score as a win condition (like the start+5000 you mention), which I expect to be more obscure to players not intimately familiar with Japanese Mahjong. So to simplify, I put a simple sudden death to avoid having a tie for 1/2, irrespective of how the split comes about and irrespective of how many points the eventual winner has. That said, I have clarified the option text and actually fixed a bug related to this.

Re 7. As discussed before, I don't share your enthusiasm ;-)

Abortive draw: as I know the rules, an abortive draw is generally considered a replay, so no rotation. riichi.wiki/Tochuu_ryuukyoku

I am marking this as implemented. Of course you are free to reopen individual points.
In any case, since you have such a vast knowledge of the game, I'd very much appreciate it if you would help move this game toward Beta at this point. Like many Alpha-games without direct Dev-involvement, there is a danger of it getting stuck, simply because not enough players can see/play it (and the dreaded training mode is forced upon everybody). Thanks regardless for all the suggestions!