I can tell you that they are not slower then one another in
principle. Quality of implementations trumps every theoretical
aspect here. LALR is usually fast even if implemented by book
but they are hard to optimize futher and quite restrictive on
"semantic extensions".
Proper CFG parsers all are liner-time anyway.
To be picky here:
The languages that can be parsed in linear time are a strict
subset of CFGs.
However I do agree, that error handling and flexibility are more
important than raw speed. I don't want to get a concrete syntax
tree full of unneeded productions, like the ones you get when you
encode precedence rules in your grammar.
