First you have to get agreement within the scientific community (not difficult), then within the worldwide legislative community (tricky), and finally get common acceptance in the general population to replace whatever traditional method they use to name or number years (damn near impossible).
When I was growing up, at school everything was BC/AD. BCE only became a "thing" to me quite recently, presumably because I don't have any exposure to discussions of history. I appreciate BC/AD is pretty irrelevant to anyone in a non-Christian environment, but BCE is just a PC re-labelling of BC and changes nothing. The datum point is still irrelevant in a non-Christian background. Nonetheless, that system of year numbering is prevalent in the Western world; I have no idea how well accepted it is in the hamlets of deepest China (I guess it is more-or-less unknown, and pretty much irrelevant to them).
To whom is the selection of the arbitrary datum relevant? It's so long ago that it is not within the direct experience of ourselves or anyone living, or even ancestors as far back as one can reasonably trace. Science has to deal with a much longer time span, and setting a datum suitable for that would lead to ridiculously large counts of years for domestic purposes, which is impractical. It is therefore only of relevance to historians, particularly those who deal with a historical period that spans the datum.
So we might as well stick with what we've got, even if it has acquired a PC label. It is mathematically inconvenient that 1BC is followed by 1AD (with no year 0, and yes, I see no reason to change to BCE/CE just to be PC), but there was no concept of negative numbers with zero occupying the space between positive and negative on a continuum, so it is no real surprise, and again it is only historians trying to work out the exact interval between events who are inconvenienced.
Astronomy has dealt with this problem by coming up with the concept of the Julian Day Number. This is a time scale where the primary unit is a day, counted from a datum sufficiently far back as to have positive values for all astronomical events of any historical interest at all. This removes any inconvenience or confusion with the missing year 0, different calendar systems (eg Julian/Gregorian, Chinese, etc etc), or the 11 days that got removed from the calendar in 1752 when the Gregorian* system was adopted in Britain. With no months (with varying numbers of days) or years (with varying numbers of days) to deal with, JD is a linear scale convenient for time calculation - it just needs a bit of conversion when presenting a JD as a calendar date (in whatever calendar you choose).
The Julian Day Number datum is Jan 1 4713BC (Julian calendar, equivalent to Nov 24 4714BC Gregorian), at noon (solar transit of the prime meridian). Time of day is then expressed as a decimal.
The current time, 10.23UT on 27th Feb 2017, is JD 2457811.93 (to two decimal places). I guess Star Trek's Star Date system must be something similar (if it has been worked out properly at all).
* The Julian calendar system has a leap year every four years, but because the orbit of the Earth does not take exactly 365.25 days on average, the result was for the occurrence of the equinoxes and solstices to drift away from the nominal 21st March/21st June/21st September/21st December - accumulating an 11 day error by 1752 (Julian calendar). The Gregorian calendar (devised by or in the reign of Pope Gregory XIII) was instigated in 1582 (and only adopted in Britain in 1752), deleted 10 days from the calendar to bring the equinoxes back in line, and modified the leap years to produce an average year of 365.2425 days - which is much closer to the actual value of the tropical year 365.242 days (ish - it drifts). This is done by eliminating the leap day from century years unless divisible by 400 (in the Julian calendar, all century years are divisible by 4 and therefore leap years). Thus, in the Gregorian calendar, 2000 was a leap year but 1900 wasn't and neither will 2100 be.
