Shall we break this down to its core, to avoid the mudslinging? Much like happened with Kiryienka, some people want to discredit a guy's former palmarès in order to manufacture a performance jump at Team Sky. Poels already had reasons to suspect him before Team Sky, and Team Sky's many shifting stories, dubious connections and rider performances make them in and of themselves a reason to suspect somebody who rides for them. It's not necessarily cut and dried (
occasional exceptions can be found) but if somebody says they find Poels suspicious because he rides for Team Sky and they don't trust that team at all, that's not a problem in and of itself.
However, it is disingenuous to say that because he had no results in the Monuments that Poels winning a Monument as a surprise outsider is automatically a reason to say he's doping - plenty of people have won Monuments unexpectedly. In recent years we've had a rider
take a monument as their first career win for goodness' sakes. Team Sky have never really shown a great aptitude for Classics tactics, because their tactics have tended to be based on a variant of their stage racing tactic (have strongest rider in race -> have strongest domestiques in race -> ride on front until everyone drops) which is harder to operate in a one-day race. It's no coincidence that Ian Stannard is probably their most successful one-day rider before today, a guy who has a very good tactical head on his shoulders and rides well when isolated.
Poels may not have a great one-day record but he's not a total schmuck either. He has good results in hilly stages of races like Tirreno-Adriatico and the Vuelta over several years. There's
stages like this that are up and down all day - the same pre-Vuelta medium mountain stage where Chris Froome lost eight minutes - and
this shows ability in a punchy finish against top names.
This is Poels doing well in the GC of a stage race through the same kind of terrain, and
this is Poels beating the likes of Valverde and Samu on a medium-sized mountaintop finish in a prestigious stage race.
Those are all before he went to Sky.
Now, is he a left-field winner of Liège-Bastogne-Liège? Sure. Is he as remarkable a winner as Maxim Iglinsky was? Maybe, maybe not. Iglinsky's previous Classics results had been mainly at races more like de Ronde, but he'd been the last man left with Contador at the 2008 Giro in the mountains on a few occasions. Is he a guy that we should be asking questions of? Well, he came from Vacansoleil, the team whose contracts catch more dopers than WADA, and he rides for a team who are, like it or not, at the forefront of doping discussion for a variety of reasons. Of course his results in the Ardennes are significantly better than before this season (although his results in Flèche had been improving year on year for a while, he'd never made a reasonable mark on LBL), but looking back at some of the stages he's performed well on in the past, that isn't necessarily because he lacked the necessary physical tools to compete in LBL because it would seem he had them; 2nd only to somebody who won LBL in that Pologne stage, 2nd only to a 2x Lombardia winner on a steep hilltop finish in that Vuelta stage, 1st ahead of a 3x Liège winner and a guy he was sprinting against today in a País Vasco stage. Clearly he lacked some racing brain, or just hadn't had the legs on the right day before. It's hard to believe that the tactical demagogues at Team Sky will have resolved the first issue, but you never know.
We also know that today's race was 9km shorter than the 2007 edition before the start, and was shortened further mid-race by an additional 5km, so making a direct comparison in race time is impossible (if you want to make the exact comparison, Poels won in a time ten minutes slower than Valverde last year. Turns out the weather was a factor. Same as why Heras' Angliru ascent in 2002 is slower than other years, strangely enough).
If you want to doubt him, that's fair enough. There are plenty of fair reasons to do so, without the need to fabricate or falsify to create additional ones.