This is what standards of proof are for. "Beyond reasonable doubt," "preponderance of the evidence," etc. You use different standards for different types of trials. In one trial you may ask, "is it more likely that this is a forged document using the default font, or that the user went out of their way to download an obscure font before its public release and specifically chose to use it?" In another trial with a different standard you may ask, "is there any possible way that this document was not forged?" The answer to each question is different, which is why we declare the standard of proof before the trial begins.
Of course, I do not know the standard in use in this country's court system.
Well, in politics as opposed to a criminal trial there's no such thing as burdens of proof, just whether a scandal has legs.
Among American (and British, Canadian etc) lawyers who advocate quantifying the concept of "reasonable doubt,"[1] the consensus figure seems to be that proof beyond a reasonable doubt means greater than 95% certainty (or, stated negatively, a less than 5% level of uncertainty). Judge Weinstein of the Eastern District of New York[2] puts the 95% number in his jury instructions, for example.
If we assume the Bayesian prior that it's at least 19x more likely that Pakistani politicians at the highest level of influence are secretly corrupt and/or evading taxes, than it is that they're secretly tech nerds who eagerly download the latest Microsoft betas and try out all the new fonts, then we have met that 95% standard. If the accused were a Microsoft executive instead of an official of one of the world's most corrupt states,[3] the equation would be different.
Of course, there might be other arguments for doubt besides "the font was available in obscure beta software." One potential argument is that it's a newer version of an older document that was typeset in a different font, which was changed to Calibri at some later point (recent MS Words might default to displaying the document in Calibri if it were sent to a computer that didn't have the original font installed, or the file type doesn't include font information in a way Word supports, or it was copied and pasted from the clipboard).
[1] Which IMO it should be; most of the clear miscarriages of justice in either direction that I know of seem like they could have been prevented or at least been mistrials if, as much as possible, we'd nudged the jury to carefully quantify the strength of both sides' evidence instead of voting emotionally.
I begin to realize the extent to which mere eccentricity (say in choosing fonts) is already deeply illegal (once you factor in how many decisions we all make that put us in a 5% box.) No wonder people herd.
Of course, I do not know the standard in use in this country's court system.
https://en.wikipedia.org/wiki/Legal_burden_of_proof