Although I do admit huge upsets can happen. My best non-rated upset was back when I was 1100, and a 2300 hung his queen on me in a queen + several pawns endgame. My best rated upset was beating a 2200 when I was 1647.
It does not mean much. If you play a lot of chess you'll eventually meet strong players when you are playing your absolute best and they are a bit off form. In OTB chess I have beaten a few IMs and significantly more FMs, but that is just a handful of results in a very long chess "career".
I have never beaten a GM in one-on-one competition, despite having reached some objectively winning positions. At my peak I was rated in the mid 2100s. If a 2100+ player cannot beat an average GM (about 2500) from a winning position then what chance does a 1300 player have?
Now, decades later, I am no more than a mid-1900s player OTB. But that is still strong enough for me to beat genuine 1300-rated players at slow time limits, without needing to exert myself at all. I need only play moderately sensibly and avoid blundering until I am given some material or some massive positional advantage. The only possible exception is a rapidly improving youngster whose published rating is lagging far behind their true playing strength.
Think about it. 2500 beats 2150 with ease from a very inferior position. Strongest player ever beaten by the 2150 in hundreds of games was an off-form 2450 IM.
Old 1900+ player beats 1300s with ease, time after time after time.
The gulf between 1300 and 2700 is vast!
No.I addressed the person who suggested the 2700 would intentionally lose games, sandbag to 1300 then beat a 2700, thus making up a whole new, absurd, scenario where the question becomes possible.
Sandbag all the way from 2700 to 1300 is a very large job, and it is psygological diffucult for a superGM to lose that much. He will also lose income and invitations when the rating drops too much. Actually I have never ever heard of GMs sandbagging.