Match or exceed v1 performance

[jhollinger]

After V1's recent perf improvements, V2 wasn't looking so great. These changes make V2 competitive again:

# Collection results (20 fields, 5 objects, 2 collections)
1,000 widgets 100x: V2 change: -59.72%
500 widgets 100x: V2 change: -60.23%
250 widgets 100x: V2 change: -60.50%
100 widgets 100x: V2 change: -61.48%
25 widgets 100x: V2 change: -62.42%
5 widgets 100x:  V2 change: -60.14%
1 widgets 100x:  V2 change: -54.12%

# Object results (20 fields, 5 objects, 2 collections)
1,000 widgets 100x: V2 change: -32.68%
500 widgets 100x: V2 change: -33.49%
250 widgets 100x: V2 change: -32.39%
100 widgets 100x: V2 change: -35.35%
25 widgets 100x: V2 change: -36.85%
5 widgets 100x:  V2 change: -54.16%
1 widgets 100x:  V2 change: -52.25%

Some major changes include:

• The DSL now builds what is essentially an AST that's evaluated on first render or reflection. Among other things, this allows use to behave like a macro within the DSL.
• The ability to use the DSL in Ruby modules, then include them in Bluerprints/views.
• around_blueprint_init can now modify blueprint options and field details.
• Lots of "legacy" extensions to ease the transition. Covers legacy if/unless options, extractors, transformers, dynamic options, and more.
• Documentation branch: jh/release-2.0-docs

[jhollinger]

@lessthanjacob This PR has really ballooned, so here are some notes:

1. Speed changes are largely isolated to serializer.rb, field_serializers.rb, and hooks.rb.
2. Changes in v2/base.rb, dsl.rb, and specification.rb (new) are largely about making the DSL build a sort of AST. This helped fix some edge cases and opened up some cool possibilities (i.e. modules).
3. Quite a few "legacy compatibility" extensions have been added under lib/blueprinter/extensions. Will help people (and agents) simplify upgrades.