Because he's not popular enough.
First round results look like
this. Second round will be between the two candidates who got the most votes on the first round, assuming nobody got over 50% (in which case there wouldn't be second round).
Assuming that everyone who voted on the first round will vote on the second round, he'd need to get over 70% of the 'homeless' votes, which is pretty much a no-can-do-scenario. Especially because I suspect there will be less votes cast on the second round; the two possible choices are both pro-EU, quite progressive candidates. One of them considerably more pro-NATO (even if he prefers to hide it) and conservative, another considerably more pro-greenish thingies (even if he's not making a big deal out of it) and liberal.
They are quite similar in many ways; neither of them is anti-EU, or patrio... errr, xenophobic, for example. That, and the lack of a left-wing candidate (Winland has had a left-leaning president for the past 30 years) on the second round will quite likely mean that plenty of people will just skip the second round altogether.
Also, there really isn't clear and obvious inheritance of votes there; our gay liberal *might* gain some votes from the left-wing voters, but there are not enough of those around this time. And our straight conservative *might* inherit some of the centrist votes, but his problem is that the two centrist candidates (who failed to get to the second round - quite likely because they split the centrist votes between them) were very, very strictly anti-EU, anti-NATO, anti-immigrants, and pro-nationalism, pro-xenophobia, and pro-racism (well, one of them isn't and the other one would never admit it... but if your party has representatives who openly mock non-Christians and openly publish fantasies about killing gays and enslaving niggas... errr... wai, yes, you haz a racist party
You also haz a party that causes great LULZ in media; not even they openly condone the actions of their more outlandish representatives, but unfortunately for them, these people 'engage mouth first, brain second, deny everything later'. Which leads to all kinds of LULZy explanations afterwards.
).
Having said that, the final results will probably be lot closer than might be apparent from the first round results; I forecast the usual 55-45 split. Still, this one should go to conservatives - unless something very unexpected happens. Not that it would matter too much; they are the two best candidates, and both of them quite good.