You didn't say what fuel you were using before you first put "hi test" in your accord. I suspect, given the improvement you observed, that you had been using "regular", and that the Honda is similar in its designed fuel needs to the Volvo.

Almost all Volvos, for a couple of decades (and with possible the exception of the S60R and V70R) covering both the RWD (red block) and F/AWD (white block) cars, have been consistently designed (at least for the US market) so that 91 is optimal, but they can still run well on as low as 87 octane or AKI, which in the US is (RON+MON)/2; in Canada, it may be different.

[Clarification: when I use the word "optimal", I mean that retardation will not occur; when I use "well", I mean that there will not be any noticeable ping or detonation because retardation will occur occasionally, when necessary to protect the engine from damage.
In older owners manuals for, e.g., 240's, it reads that 91 is "optimal" and 87 is "minimal"; in later manuals for, e.g., S60's, it reads that 91 is "recommended" and 87 is "minimum".]

The difference in these AKI numbers is the occasional intervention by the knock sensor for ignition timing retardation. That is, with at least 91 octane, there should be no appreciable instances of, or need for, spark retardation. With lower AKI, retardation will occur occasionally (depending on circumstances such as weather, load, and driving behavior). Of course, any retardation reduces fuel mileage, as well as power output.

Thus, depending on how you drive and when and where, it's quite possible (and predictable) to see an improvement in mileage as you described.