I still don't understand how could user be affected by this discovery even without updating to the latest firmware?
This is really an edge vulnerability, and most likely the majority of users keeping even large amounts by personal standards are not at serious risk of the loss of coin.
The specific type of display the Trezor One uses will consume different amounts of power depending on how many pixels are displayed on the screen. The Trezor One display will also start to display one line of pixels at a time with each subsequent line being displayed fractions of a second after the prior line. This means someone monitoring the power consumption of your Trezor One can determine how many pixels each line your Trezor One is dissplaying at a time. An attack could use this information to reasonably guess what is being displayed on your Trezor's screen.
If you were creating a new seed with your Trezor One, an attacker could learn the seed words, and the position of each word that the Trezor One displays. An attacker could also know which row each number is displayed on your Trezor One when displaying the numbers when you enter your PIN; this will allow an attacker to learn your PIN if they monitor your Trezor One's power consumption and monitor your computer after you enter your PIN multiple times.
In order for this attack to be successful, an attacker must have physical access to your computer, and they must install specialized equipment in your computer without you noticing. The attacker must compromise your computer *before* you use your Trezor One on the computer, and cannot learn any information after the fact.
This attack would be specifically targeted to its potential victims. The vulnerability has already been patched with new firmware that instructs the Trezor One to display additoinal random pixels that makes this attack vector moot.