Exercise 2: Matched Capacities — Possible Solution
====================================================================

SETUP
------------------------------
    REVIEW_RATE raised from 2 to 4, matching BUILD_RATE

RESULT
------------------------------
    Day  1: built=4,  reviewed=4,  waiting=0
    Day  2: built=8,  reviewed=8,  waiting=0
    Day  3: built=12, reviewed=12, waiting=0
    ...
    Day 10: built=40, reviewed=40, waiting=0

"Waiting for review" stays at exactly 0 every single day - no backlog
ever forms.

WHAT THIS REVEALS ABOUT THE CFD SIGNAL
------------------------------
A cumulative flow diagram's own bottleneck signal is entirely relative
- it's not about any single stage's absolute speed, it's about whether
one stage's rate exceeds its neighbor's. The chapter's own mismatched
rates (4 vs. 2) produced a visibly widening gap; identical rates (4 vs.
4) produce a perfectly flat gap, regardless of whether that shared rate
is fast or slow in absolute terms. A team watching only Build's own
output number (still climbing at 4/day, same as before) would see no
difference between the two scenarios - the CFD's own "waiting" band is
what actually distinguishes a healthy flow from a growing bottleneck,
which is exactly why it's tracked as its own metric rather than
inferred from either stage's throughput alone.