With a very big probability, it's a conspiracy theory - and even if someone might try, it doesn't yet mean it actually works
Sure, the juries might base their votes on the odds, but it is very difficult to influence odds when the market is this big.
S-markets shows more than 1.5 million pounds of bets on the winner already (you can check that
on this link), and that is just one betting market - the eurovisionworld.com compares about ten of them, and I am sure there are many others that don't even make it to this aggregating website. So, let's guesstimate (with a big dose of pessimism) the total amounts of bets made at the moment at 5 million. And let's not forget you actually need to be high in the odds in May when this number explodes further compared to now, because most people are casuals who make their bets once the event is just around the corner. Once again, let's be pessimistic and say the total amount people bet during the entire season is 20 million €... Based on that, to make any meaningful lasting impact, I estimate you would have to bet higher tens of thousands of € at the very minimum throughout the season, possibly a lot more, with a very uncertain outcome at that.
It's much easier to just take that money and spend it on adds / a better staging. Even manipulating the televote of some smaller / less interested countries seems more feasible and makes more sense.