sql - Does the size of the data stored in the record affect the Clustered index performance? -

i have 3 tables.

1. primary fact table (facttable)
2. a table containing int field + other 'n' varchar(50) fields(table001)
3. a table containing int field + other 'n' varchar(max) fields(table002)

records in each table: #5 million rows

query:

select * facttable f (nolock) inner join table001 t001 (nolock) on f.id = t001.id inner join table002 t002 (nolock) on t002.id = f.id f.datefield = '2014-02-01' 

all tables have clustered index on field 'id'(which used in joining tables i.e., foreign key)

the clustered indices not fragmented(fragmentation percentage < 5%) on tables

the case when join 3 tables, seeing in execution plan cost of joining table002 higher table001. there reason behavior?

link xml execution plan: https://onedrive.live.com/redir?resid=7e436ae9d73999b0%21120

i couldn't paste xml content because of size restriction.

another link : http://www.expinos.in/executionplan.sqlplan

the size of columns in row, not in index, not effect performance of containing index itself.

however,

large non-indexed columns effect performance of table based operations. you'll less rows on page, when scanning, you'll have read more pages.

large non-indexed columns included in query have obvious direct effect on query performance. you'll have read bigger values disk/memory, although optimiser try late.

the size of columns in index effects performance of index.

the size of columns in clustered index effect performance of indecies on table

---

is easier compare 1 byte or 10000? easier hash 1 byte or 10000? easier read 1 byte disk/memory or 10000?