Correcting the calendar and time conventions

Noting the odd convention of the leap year, multi-base (non-decimal) definitions of time units with seemingly arbitrary multiples which aren't always integers, and discrepancies between the conventional time for the average year of 31,557,600 seconds per year and the correct time per year of 31,556,925 seconds per year, one might resolve these very small temporal discrepancies in the time conventions by restructuring the standards of time so that one year measures the period of revolution of the earth around the sun with little need for corrections (every four years) and that each unit, like the metric system is a multiple of ten as opposed to the 60s per minute, 60 minutes per hour, 24 hours per day, 7 days per week, 4.4 weeks (constantly varying, but never an integer number, between 4-5) per month, 12 months per year, 52 weeks per year, etc.

With 31,556,925 seconds per year (365.25 days) in 2000 and 31,556,981 seconds per year (365.25 days) in 2100, the latter standard will be good for the next 200 years as an average to the year 2200. With 365 days per year, an extra 57.48 seconds is added to every day, yielding an average day of 86,458 seconds (86,400 currently). A new second can be defined as 0.86458 old seconds to yield 100,000 seconds in a day, which can be divided into new decimal based time units of 10 hours per day, 100 minutes per hour and 100 seconds per hour.

Given 365 days, we could follow the lunar cycle, with 13 months in the year, where each of the first twelve months has 28 days and the thirteenth month has 29 days. Or if we have an aversion to the number thirteen or to changing the number of months in a year, then we could define 5 days per week would yield 73 weeks, thus 12 months per year would allow 6 weeks per month for 11 of those months, with 7 weeks in the 12th month. These considerations have been made with the restrictions that the length of the day must remain relatively the same and that the year must extended to maintain itself from year to year without any need for leap year corrections, and that all units of time should been even multiples of each other with the preference of base ten numbers and a secondary preference of numbers similar to standard conventions when the primary preference is not ideally applicable.