Disconnecting the cable will have no affect.
Even so, using a low reading ohmmeter (spec: 4-24v d.c. with 200mA short circuit current) the reading you get will never calculate to the same value you read from the meter. An R1+R2 value for a standard lighting circuit will be around that value you got.
I take it you deducted the lead resistance first though right? EDIT: Read you nulled them.
But, even when you measure Zs it never works out the same as the calculated method either. Why? Because of the earth path taken during the test will have parallel routes via exposed and extraneous conductive parts to effectively improve the earth fault loop resistance. Couple that with the fact that the temperature at the resisitances given in the On-Site Guide are at 20 °C, and fluctuations in supply voltage, you will spend many, many hours scratching your head to try and balance the figures.
Ideal test situation to get the measured and calculated to match:
@20 °C
Unbroken conductors
Completely independent from the equipotential zone
Even then.. it won't match
The is no means to fail verification of a Method 2 (R1+R2) test anyway. The circuit will achieve shock protection if the Zs value is equal or less than the On-site Guide value or equal or less than 0.8x of the values given in BS7671:2008
Here's one for you though.
Why don't we deduct the lead resistance from our Zs tests?