Excuse me, I shouldn't have laughed. Ahem. Mercury has the highest delta-v requirement amongst the planets.
Which would have been a great response, had I written:
Venus and Mercury have the lowest delta-V requirements
Which is, of course, NOT what I wrote. What I actually wrote was:
Venus and Mercury have more frequent launch windows.
With Mercury you can't save energy by aerocapture, but you can make very frequent flybys of Venus and Mercury to lose velocity - and making observations of both Venus and Mercury during each flyby. Messenger, for example, was launched in August 2004. It did an Earth flyby 1 year later, a Venus flyby 12 months later, then a Mercury flyby 15 months later, then 3 years later, orbital insertion. Now, that was a long time to orbital insertion, but not a long time before the collection of first data - it did minor data collection at Earth, more at Venus, and more during its first Mercury flyby. But more to the point, it was done with a tiny delta-V budget: 316m/s from Earth escape to Mercury insertion and 862m/s for the orbital insertion itself. Mars takes about 390 m/s from Earth escape to Mars transfer, 670 m/s for Mars transfer to Mars capture (if you can't aerocapture), and entering a similar orbit on Mars would take somewhere in the ballpark of 2000 m/s (5000 m/s to low orbit) (again, possibly reduced if you can aerobrake). While it's technically possible to use an Earth gravity assist to Mars, it takes much longer.
Now, of course, one could have used a more Mars-like delta-V to get to Mercury much faster, although due to the complexities of gravitational assists I can't work out readily here just how long it would take. But for Venus it's easy: 280 m/s from Earth to Venus transfer and then... well, aerocapture is pretty much a given if you want to (far easier than on Mars), but if you don't want to do it, then it's another 360 m/s to high orbit, then 2900 m/s to low orbit. But of course, Venus is a very easy body to aerocapture at, so it should be expected. Short transfer times to Venus, little energy to get there, easy to brake at, and frequent launch windows. How is that not more "probe-able"?