You haven't really defined hi and lo. But I'd say anything is possible.
Concentrations depend on generation rate profile over time and removal rate profile over time, among other things.
One thing to note is Ostwald solubility coeff of H2 (0.04) is much lower than C2H2 (1.0). For equilibrium concentrations between oil and gas, that means a much higher fraction of H2 migrates into the gas. In the case where the transformer is breathing, we might conclude a lot more H2 is lost as a result of breating than C2H2 is lost. (although Ostwald coeff isn't the whole story since it relates equilibriums but not time to reach equilibrium).
So let's say you had some type of fault event a long time ago and generated lots of H2 and C2H2 and stopped generating. After that, the transformer breathers for a long time and loses a lot of H2 but not much C2H2. Then you come along and sample and see much higher C2H2/H2 ratio than you would have seen initially.
“Anything’s possible” seems like a safe answer. To get better input, (there are plenty here that can provide valuable input), would help to know more details.
What do the other gases look like?
How has the pattern changed over time?
I assume this is main tank (not LTC)?
What type of preservation system (N2 blanket with bottle/regulator/relief, conservator, breather etc).
