The Huawei HG612 uses a Broadcom chipset and I don't think many would take issue with the recommendation to use a Broadcom based device on a Huawei cabinet (which also use Broadcom chipsets). However, there are some examples where a HG612 has been found to underperform compared to more modern Broadcom based devices, also these inexpensive devices don't necessarily use top grade components, especially when it comes to capacitors, so their performance can deteriorate over time.
I'm still using a HG612 revision 3B on my Huawei cabinet line and have no problems with it - but there might be better choices for not much more money than buying a HG612 off eBay (these devices are not available new - they were a special product for Openreach).
I don't believe Openreach have made a formal statement about VDSL2 vectoring, nor are they likely to do so, but it seems that they have made a firm decision not to roll out VDSL2 vectoring in a widespread fashion. Vectoring is expensive to roll out and this tends to dwarf the returns, though it has been or will be deployed in some cabinets if it proves cost-effective in helping to meet BDUK coverage targets or where LR-VDSL will be offered (in which case vectoring is mandatory to avoid horrendous crosstalk between the Profile 17a VDSL2 signals of the ordinary FTTC service and the Profile 8b VDSL2 of LR-VDSL originating from the same cabinet). There might be further limited deployments to help meet a future broadband USO falling on Openreach (other than in cabinets that are to offer LR-VDSL) or to work round a specific local issue such as poorly performing D side cables. In other words, Openreach appear to be viewing vectoring as a cost-effective way of improving the speed on some longer lines, but not a cost-effective way of offering higher speeds on short lines.
Without vectoring - or an equally dead-end move to something like Profile 30a (which is likely to be even more costly and have more compatibility issues than deploying vectoring) - there seems little prospect of Openreach offering more than 80/20 over VDSL2. Rather than making heavy investment in dead-end improvements to VDSL2, it seems that Openreach have decided to invest in G.FAST deployment and more FTTP in the pursuit of faster speeds.
If all you are after is a modem bridge, I'd spend the least money you can for something of reasonable quality and view the resulting item as disposable. Trying to future proof an inexpensive item makes little sense; you'll land up paying additional money for something that might turn out not to be future proof after all. There is no guarantee than any G.FAST equipment bought today will turn out to work at all with the commercial Openreach G.FAST service when it arrives, let alone work well with it.
It is a little more difficult if you intend to buy an expensive router but, so long as the router supports Ethernet WAN, it can be used with any current or near future technology when using the provider's supplied equipment or a relatively inexpensive modem bridge.