I am trying to answer this Phys.SE question but got an unreasonable answer:
In making laboratory measurements of $g$, how precise does one have to be to detect diurnal variations due to the moon's gravity? [Clarifications omitted.]
I figured I can calculate the change in GPE for a test mass
$$GPE_{earth} - GPE_{moon}$$ at apogee and perigee, and take a ratio to see how many significant figures are affected.
Where $d$ is lunar distance,
$$\frac{m(g_{earth}R_{earth}-g_{moon}d_{min})}{m(g_{earth}R_{earth}-g_{moon}d_{max})} = \frac{m_{earth}R_{earth}-m_{moon}d_{min}}{m_{earth}R_{earth}-m_{moon}d_{max}}$$
I asked Wolfram Alpha to calculate this and got a ratio of about 1.4, which seems unreasonable.
What has gone wrong?