Actually the Mars surface is easier to get to in terms of delta v than it's moons. That's because the moons orbit so close to Mars, so are hard to as far as the delta v budget is concerned. While with the Mars surface, you can slow down to a stop using aerobraking.
Robert Zubrin puts this as a reason for targeting the surface rather than its moons. It's far harder to get back from the surface though of course, which is why there is all this talk about one way missions to the Mars surface.
However there's another solution, better than Phobos or Deimos for an early mission - that's to go into the Molniya style orbit, which is close to a Mars capture orbit.
This is roughly as easy to get to as the surface of Mars in terms of delta v - but far easier to return to Earth - almost no delta v needed to get back to Earth at all, less than 1 km/ second delta v needed for the return journey, that's just for the maneuver needed to send it on a trajectory back to Earth (no need to decelerate for return to Earth as you use the atmosphere for that)..
It's a spectacular orbit, gets you close to the surface of Mars every 12 hours, and you could use that for close up telepresence operation of robots on the surface. As far as science value is concerned that's a big advantage - telerobots on the surface don't need life support or to carry air, drink, and food around with them or living space or spacesuits or toilet facilities - and there are no astronaut lives endangered - they can do things like climb cliffs, even fly (as they can be lightweight) and so on - and can be deployed to many different places on the surface. And you don't need to develop a human rated lander.
The biggest plus of all though is that surface telerobots can be sterilized and so can be sent to sensitive areas of Mars that might host present day life - while humans can't be. Also human ships could crash on Mars which arguably greatly increases the chance of irreversible contamination with Earth life, while telerobots can be sterilized so that that is not an issue.
For all those reasons I think myself that the first missions to Mars will be like this - that is if we send humans there at all. Humans would greatly speed up discovery and exploration of Mars at present stage of technology if exploring it from orbit like this. But autonomous robots are getting better also - and a fast capable autonomous robot on Mars could also help a lot too.
Depends how soon the human missions get underway - if they are delayed another two or three decades we might perhaps get autonomous rovers able to take over many of their functions, especially autonomous driving and routine experiments - who knows.
Anyway - this is the Molniya style orbit around Mars, as you see a spectacular orbit, nearly as easy to get to as the surface in terms of delta v, almost no delta v to return to Earth, and opens up both sides of Mars, on sunny side, to close up telepresence work,
Here is a video animation of what that orbit is like (animation done in Orbiter with a futuristic spacecraft, some time will do a more realistic animation with present day type tethered Dragon or some such, but the orbit is correct):
Then - at an earlier stage, an even easier human mission is Robert Zubrin's Robert Zubrin's Double Athena Flyby - mission. It never actually orbits Mars but instead does two flybys. One of them puts it into an orbit around the sun similar to Mars, it then co-orbits the sun with Mars - slightly outside of Mars but not far from it, until it comes close enough again to Mars for its second flyby roughly half a Mars orbit later (so one Earth year later) -which then returns it to Earth.
It is a "free return" mission - once you are on your way, you don't need to use any extra fuel to get back to Earth except minor course corrections. And you get a couple of flybys in the one mission, each with some hours of close up telepresence operation of telerobots on Mars, and for days you are close enough to Mars to do useful telerobotics - and for a whole year, you are so close that it's a huge improvement over operation of the Mars surface rovers from Earth.
Deimos does have advantages also though. It's nearly synchronous, so you have several hours a day of operation of rovers over any one spot on Mars. It also has craters at its poles which are the only places in permanent darkness in the inner solar system that we know of apart from the poles of Mercury and the Moon. This could make them of scientific interest - it also provides protection from solar flares and passive cooling of the spacecraft. Though of course if you had a spacecraft in one of those craters, you'd need to get power from solar panels erected outside of it.
Also from its spectragraphic type, astronomers think there is a good chance, though not certain, that Deimos may well have supplies of ice underground which you could use for rocket fuel - if so that could negate many of the disadvantages of the greater delta v you need to use to get there.
Phobos is of special scientific interest because it is so close to Mars and they reckon that it has a significant percentage of Mars material in its regolith, from impacts on Mars throughout its history. This makes it a potential mission for a sample return of Mars material - and one that does not run the same risks as a surface sample of potentially including still viable XNA based lifeforms with unknown effects on the Earth's environment. It's already sterilized by cosmic radiation - and if you wanted to be sure (I think maybe advisable for first sample returns) could zap it with a bit more gamma radiation and have almost no effect on its science value but makes sure there can't be life on it.