page 1

s/nearly, as fast/nearly as fast/

page 2

s/It doesn't try to compile quickly for the majority of code/
/It doesn't try to quickly compile the majority of code/

s/often a one/often one/

s/fast hopefully common case/fast (hopefully common) case/

s/the system the developed system/the developed system/

In the 3rd paragraph of 2.1 you say something like 'may have been'
referring to the object memory design.  I suspect a quick review
of the Back to the Future document may help to remove the speculative
tone:  ftp://st.cs.uiuc.edu/Smalltalk/Squeak/docs/OOPSLA.Squeak.html

ASK ABOUT DETAGGING

page 3

s/Modern CPUs are very fast but performance isn't simple/
/Modern CPUs are very fast but making efficent use of this speed isn't simple/

What does 'normally optimised' mean?

'taken jump'?

s/Appel's "Modern compiler implementation in C" combined with the Urs Holzles' work in self/Appel's  _Modern Compiler Implementation in C" combined with Urs Holzles' work in Self/  Book titles should be underlined or italicized.

page 7

Is SSA spelled out anywhere?

page 8

s/eliminate all the overhead however for/eliminate all the overhead; however for/

page 12

s/VisualWorks which is/VisualWorks is/

s/than VisualWork for the benchmark\./than VisualWorks./

s/C is slow either a memory/C is slow due to either a memory/

page 13

s/be speed up/be sped up/

s/of both do not/of both does not/

s/Such a optimiser/Such an optimiser/

In general you use 'should' a lot which makes it sound rather speculative.
You might want to go through the text once and try to remove as many of
those as possible.

A bibliography would be nice even if it only lists the details for Appel and
a link or two for useful papers from Holzle.