Refactor HitResult generation to make ruleset-specific behavior more explicit

[Givikap120]

Before you had function GenerateHitResults(double accuracy, IBeatmap beatmap, int countMiss, int? countMeh, int? countGood), where you pass countMiss, countMeh and countGood.
So by function signature you may assume that it's using those 3 judgements to generate hitresults.
In reality - osu also used largeTickMisses and sliderTailMisses from class itself, mania used oks and greats from class, when taiko just ignored countMeh because it doesn't exist in taiko. What is quite confusing and unintuitive.

This PR removes those params from basic GenerateHitResults, so it would be intuitive that all values are taken from attributes of the class.
And then this function delegates calculation to ruleset-specific static function that takes correct parameters.

Prerequisite for #232

[Givikap120]

This PR, similar to #232, seems to complicate code at no gain. Nothing is wrong with the current state, and imo splitting this up is unnecessary

I explain the gain in detail in PRs description. Please respond more precisely about what is trying to be addressed instead of "PR is bad".

[smoogipoo]

@Givikap120 you're going to need to fix this PR.

[smoogipoo]

Will also await on further refactoring as per https://discord.com/channels/188630481301012481/380598781432823815/1307180578906312724 before merging.

[minisbett]

Will also await on further refactoring as per https://discord.com/channels/188630481301012481/380598781432823815/1307180578906312724 before merging.

This PR is not the one requiring refactoring I believe, this PR should be fine as-is. The accuracy calculation PR does though.

[Givikap120]

@smoogipoo I fixed merge conflicts here